Strictures of the small intestine, not obviously due to tuberculosis, have been recognized since at least the sixteenth century. In the late nineteenth century, amongst the case histories of patients with supposed ulcerative colitis, there are those who also had gross ileal involvement. In 1913 Dalziel recognized that ileal inflammation could occur as a disease entity. He described the macroscopic appearances at surgery and the histological features of a chronic granulomatous inflammation predominant in the submucosa, which he distinguished from tuberculosis. However, little attention was given to this description until Crohn, Ginsberg, and Oppenheimer published their classical description of ‘regional ileitis’ in 1932. Later, in the 1930s, it became apparent that the disease often affected the right colon in continuity with terminal ileal disease, but primary disease of the colon was not accepted until the description of Lockhart-Mummery and Morson in 1960.

The presence of disease in the colon made the term regional ileitis or terminal ileitis inappropriate. Subsequently, regional enteritis or granulomatous enteritis has been used, but the eponymous term ‘Crohn's disease’ is preferable, since it does not imply an anatomical site nor the presence of granulomata.

Crohn's disease may affect any part of the alimentary canal from mouth to anus, although ileal, ileocolonic, or colonic involvement are the most common patterns. The disease is characteristically discontinuous, or in other words, the disease may affect a number of segments with relatively normal intestinal mucosa in between. The inflammation is often transmural but is predominantly submucosal and is associated with a chronic inflammatory infiltrate, fissures, ulcers, and fibrosis, which may lead to stricturing. Granulomata, with giant cell formation, are the hallmark of the disease but are not always present.

Crohn's disease has a worldwide distribution, although there is considerable variation in its actual incidence. In northern Europe and the United States of America, the annual incidence is about 3 to 5 per 100 000 of the population, with a prevalence of about 40 to 60 per 100 000. In Japan the disease appears to be about 10 times less common, and it is rarer in developing countries. However, diagnostic difficulties are considerable in those parts of the world where intestinal tuberculosis is still common.

The incidence of the disease in Britain, Scandinavia, and Israel increased during the 1950s and 1960s (Table 1) 313, but it may now have reached a plateau. The incidence of the disease may be increasing in children. Ethnic differences also influence the incidence, with particularly high incidence rates in Ashkenazi Jews and a low incidence in Americans of African descent. Some series show a slight female preponderance, but this is not marked. The peak age of onset is in young adults (20–40 years of age), but the disease may present at virtually any age.

The pattern of the disease appears to be similar wherever it occurs in the world, although extraintestinal manifestations are probably more common in Caucasians. Ileal disease is particularly interesting in Japan, since it is often associated with very extensive longitudinal ulcers. These can be seen in Western patients but they are rare.

The familial incidence of Crohn's disease varies widely between series, but is probably 10 to 15 per cent. First-degree relatives are mainly affected and may have Crohn's disease or ulcerative colitis. Nevertheless, it has not been possible to describe the inheritance in classical mendelian terms. There is 20 to 30 per cent discordance in monozygotic twins. It is exceptionally rare for both husband and wife to be affected. No clear association with any of the HLA antigens has been found, although in Japan there is a weak association with HLA-DR2.

The presence of granulomata suggested a mycobacterial aetiology to the clinicians of the 1930s, but no evidence for such an organism could be found.

However, in 1984 isolates of Mycobacterium paratuberculosis from Crohn's tissue were reported and the isolated organisms induced intestinal inflammation when given, by mouth, to new-born goats. Since then, several more isolates have been made in different countries and using DNA probes, most of these have been shown to be identical with M. paratuberculosis. Nevertheless, the significance of these isolates remains very doubtful, since less than 10 per cent of tissues cultured have proved positive. Secondly, the clinical response of Crohn's disease to antimycobacterial therapy has not been overwhelmingly successful and, in controlled trials, no effect has been seen. At the present time, several groups are attempting to identify M. paratuberculosis DNA in tissues, using the polymerase chain reaction to amplify the DNA. Despite the extreme sensitivity of this technique, findings have been essentially negative.

Many other infectious agents have been implicated (viruses, cell-wall-deficient pseudomonads) but no convincing evidence for them has accumulated. However, several lines of evidence suggest that luminal factors may be important. Active disease often responds to the use of elemental diets, which appear to be as effective as corticosteroids. Colonic disease, resistant to medical treatment, frequently subsides following defunctioning of the colon by means of a split ileostomy. Challenging the defunctioned colon with ileostomy effluent will trigger an inflammatory response, although this is not seen if an ultrafiltrate of the effluent is used. Thus, luminal bacteria may have some influence on the course of the disease, although the effects of elemental diets and defunctioning may equally well implicate a biochemical or metabolic alteration in the luminal environment.

Many studies have now shown that patients with Crohn's disease are more likely to be smokers, and this is in sharp contrast with ulcerative colitis, which is found more often in non-smokers. Thus smoking confers a relative risk of 3 to 4 for Crohn's disease. The reasons for this are completely obscure. Smoking is known to affect the synthesis of colonic mucus and mucosal blood flow. It also effects arachidonic acid metabolism and some immunological effector mechanisms. However, whether any of these mechanisms can explain the opposite associations between smoking and ulcerative colitis or Crohn's disease remains speculative.

There may be an increased incidence of Crohn's disease in young women receiving oral contraceptives. However, the association, if it exists, is weak and largely disappears when the data are adjusted for smoking habits and social class.

There is considerable evidence to suggest that immunological mechanisms play a role in the pathogenesis of mucosal inflammation. Thus, intestinal macrophages and T cells are activated within the inflamed intestine and release a wide variety of cytokines as well as inflammatory mediators. These soluble mediators have profound local and systemic effects, which include altering epithelial cell permeability and ion flux, initiating fibrosis by activating fibroblasts, damaging endothelium resulting in local ischaemia, and stimulating an acute-phase response (Table 2) 314.

The more difficult question is whether Crohn's disease represents a specific immunological disturbance. For example, does the chronic inflammatory nature of the disease reflect an abnormality in the regulation of the mucosal immune response? So far, evidence to support such a hypothesis is lacking. Assays for T-cell suppressor or helper function, whether using peripheral blood or mucosal lymphocyte populations, have produced variable results. However, recent data suggest that patients with Crohn's disease, in common with patients with ulcerative colitis, may have an impaired ability to induce suppressor cells to specific antigens. This inability is present in patients whose disease is in remission and on no therapy, suggesting that there may, indeed, be an underlying problem in immunoregulation.

Deficient immunoregulation would allow an immune response in the intestinal mucosa to proceed unchecked and might explain why these patients exhibit humoral and cellular immune responses to a wide variety of luminal and epithelial antigens. A particularly interesting finding is that, although patients with active ulcerative colitis or Crohn's disease may show a rise in total serum IgG concentration, the patients with Crohn's disease show a predominant rise in the IgG2 subclass whereas those with ulcerative colitis have a predominant increase in IgG1. This difference could indicate that different antigens are involved in each disease, since protein antigens are associated with an IgG1 response whereas carbohydrate or bacterial antigens stimulate an IgG2 response. Equally, it could represent a genetic difference between the two diseases.

Crohn's disease affects the small intestine in at least 70 per cent of all patients with the disease. Of these, about half have ileal involvement alone and the rest will have associated colonic disease—usually affecting the right colon. However, 30 per cent of patients with ileal disease alone will show a proctitis on sigmoidoscopy.

Macroscopically, the affected segment of small intestine is thickened due to oedema and fibrosis, which can be severe enough to cause a stricture. The mucosa is inflamed and may have a ‘cobblestone’ appearance, due to a combination of submucosal oedema and fissuring ulcers (Fig. 1) 956. In mild disease, small aphthoid ulcers may be visible on the mucosa. In more severe disease, these ulcers become larger and coalesce. Ulceration is often serpiginous, especially on the mesenteric border. On the serosal surface, Crohn's disease may have a fairly characteristic appearance, with creeping fat and marked vascularity.

The predominant histological feature is a chronic inflammatory infiltrate of lymphocytes and plasma cells in the lamina propria and submucosa (Fig. 2) 957. There is usually shortening of villous height with increased crypt depth, and there may be evidence of ‘pyloric metaplasia’.

This last finding has been investigated recently and represents a novel lineage of cells budding from the adjacent crypts. This appears to be part of a normal mucosal repair mechanism and may develop under the influence of epidermal growth factor.

When the Crohn's disease is active there is also an increase in neutrophils and eosinophils, and there is usually ulceration of the surface epithelium. One characteristic feature is the fissure ulcer, which often penetrates deep into the mucosa (Fig. 3) 958. These fissures are lined by macrophages and lymphocytes and may be associated with granulomata and multinucleated giant cells (Fig. 4) 959. Fibrosis is another prominent feature, especially in the submucosa. The inflammation of Crohn's disease is often transmural and therefore it is common to see chronic inflammatory changes with granuloma formation on the serosal surface.

Lymphoid follicles are prominent and there may be small, localized areas of ulceration in the overlying epithelium. It has been postulated that these lesions represent the earlier lesion of Crohn's disease, but this remains a speculative, though attractive, hypothesis.

Granulomata occur with variable frequency, partly depending on the number of sections examined and the time spent searching for them. However, they are probably present in no more than 65 per cent of patients with ileal disease, and some series quote only 30 per cent. Certainly there appear to be fewer granulomata in ileal disease than in colonic Crohn's disease, the frequency increasing as the disease becomes more distal. Granulomata are often associated with blood vessels or lymphatics, and some recent experiments, using operative specimens injected with intra-arterial resins, have shown that the granulomata are actually within the walls of arterioles. Although the presence of granulomata has been claimed to indicate a good prognosis, subsequent studies have not confirmed this.

Neuronal hypertrophy is another histological feature that is frequently present. Nerve endings, in particular, become prominent and stain strongly for vasoactive intestinal polypeptide, but whether this is specific for Crohn's disease, or whether it is merely associated with the neural hypertrophy known to occur proximal to obstructive lesions, is not known.

The most common symptoms of small intestinal Crohn's disease are diarrhoea (90 per cent), abdominal pain (55 per cent), anorexia, nausea, and weight loss (22 per cent). Malaise, lassitude, and the symptoms of anaemia are frequently present.

The diarrhoea usually consists of the passage of frequent loose or watery stools. Nocturnal diarrhoea may occur. Frank blood in the stool is not usually seen, although small amounts of blood may be passed in patients with an accompanying proctitis. A few patients may have a steatorrhoea but the passage of pale, bulky stools which are difficult to flush away is uncommon.

The pain may either be a dull ache, often located to the right iliac fossa, or a periumbilical colicky pain. Both types of pain can radiate into the back.

Some patients present with general symptoms of vague ill health, with few, if any, abdominal symptoms; this usually leads to a considerable delay in diagnosis and is seen particularly in children, in whom failure to thrive or growth failure are common presentations.

The symptoms of small intestinal Crohn's disease may be caused by active inflammation, intestinal obstruction, or both. Patients with previous ileal resections may have a watery diarrhoea due to malabsorption of bile salts, or their symptoms of anaemia may be due to vitamin B&sub1;&sub2; deficiency. Partial obstruction from stricturing disease may cause bacterial overgrowth, which may cause steatorrhoea, bloating, excess wind, weight loss, and a general malaise.

Thus, the symptomatology of Crohn's disease is complex, and an awareness of the differing causes of symptoms is important in making an accurate clinical assessment.

Many patients look well, are well nourished, and have few abnormal physical signs. However, evidence of weight loss, anaemia, and signs of iron deficiency and clubbing of the nails are common. Peripheral oedema may occur if hypoalbuminaemia is present. In very sick patients, there is often marked cachexia, proximal myopathy (from hypokalaemia or vitamin D deficiency), easy bruising, and pigmentation.

Abdominal examination may be normal, but right iliac fossa tenderness is common. Thickened loops of ileum or a frank abdominal mass may be palpable. Visible peristalsis usually denotes obstruction but, in cachectic patients, may just be normal small-bowel movement. Perianal disease, which includes fleshy skin tags, fissures, or fistulae, can occur in up to 14 per cent of patients with ileal disease. Rectal examination may demonstrate tender pelvic loops of inflamed intestine.

Although the majority of patients complain of symptoms which have been present for weeks or months, a few (less than 10 per cent) present with a more acute history. This often consists of right iliac fossa pain, fever, and malaise over a few days. These patients are frequently diagnosed as having acute appendicitis, and the true diagnosis is only revealed at operation. Nevertheless, a careful history often reveals more long-standing symptoms, and investigation often shows a mild anaemia or a lowish albumin. Thus, however classical the clinical picture for acute appendicitis appears to be, a full history, together with documentation that signs of chronic disease are absent, documentation of a normal anus, and a normal haemoglobin, should be essential before proceeding to appendicectomy.

The diagnosis of small intestinal Crohn's disease is made by barium contrast radiology. When available, the technique of enteroclysis (small-bowel enema) should be used, as this allows much better definition of the mucosal pattern than a conventional barium meal and follow through. Thus, scattered aphthoid ulcers are much more likely to be seen on a small-bowel enema, and it also allows a much better assessment of strictures. More advanced disease will be associated with the appearance of fissure ulcers, mucosal oedema, and narrowing of the intestinal lumen (Figs 5, 6) 960,961. Fistulae to other loops of bowel or to other organs may be visualized (Figs 7, 8) 962,963.

When a small-bowel enema confirms Crohn's disease, a full examination of the colon should always be made. It is usually sufficient to perform a sigmoidoscopy with rectal biopsy and a double-contrast barium enema. Nevertheless, the barium study will miss scattered aphthoid lesions in the colon and so, for a full assessment, colonoscopy should be performed. The importance of taking biopsy specimens cannot be overemphasized since focal inflammation and even granulomata can be seen histologically in 25 to 30 per cent of patients in whom the macroscopic appearance of the mucosa is normal. Colonoscopy may also allow terminal ileal biopsy specimens to be obtained, which can be helpful in cases where the radiological appearances are suggestive but not diagnostic.

White-cell scans (using indium- or technetium-labelled granulocytes) may also be helpful in determining the extent of the disease (Fig. 9) 964. However, their major use is in assessing patients where symptoms are more severe than would be expected from the barium radiological appearances and in excluding local perforation with abscess formation.

Computed tomography (CT) scans, especially following contrast medium, may demonstrate thickening of the intestinal wall, but they are of low specificity. CT scans and magnetic resonance imaging can be very useful in the assessment of fistulae or pelvic abscesses.

As already mentioned, an acute ileitis frequently mimics an acute appendicitis. Nevertheless, only a small proportion of patients with an acute ileitis have Crohn's disease, on the grounds that only about 10 per cent will have recurrent disease on prolonged follow-up. The remainder are assumed to have an infective cause, and Yersinia enterocolitica or Y. pseudotuberculosis account for a few. These organisms can be isolated from stool cultures, but a rise in serum antibody titre is the most reliable method of making the diagnosis.

For more chronic disease, the differential diagnosis includes lymphoma, Behçet's disease, tuberculosis, actinomycosis, and tumours (adenocarcinoma, carcinoid). A small-intestinal lymphoma may be difficult to distinguish from Crohn's disease, except following laparotomy. The presence of focal colonic disease at colonoscopy, and the appearances on an abdominal CT scan, may be helpful. Behçet's disease of the small intestine is primarily a vasculitic process and virtually always occurs when there is evidence of disease elsewhere (for example, ocular or genital ulceration, deep vein thrombosis, neurological involvement). Ileal perforation is not uncommon in Behçet's disease but is very rare in Crohn's disease of the small intestine. Radiation enteritis may also mimic Crohn's disease, but a history and radiological appearances make the diagnosis clear.

The most difficult differential diagnosis is intestinal tuberculosis when Crohn's disease is suspected in patients at high risk from tuberculosis (for example, Asian patients). Stool culture and serum antibody titres are unhelpful. Radiologically, tuberculosis often affects a very short segment of the ileum and is not usually associated with cobblestoning or asymmetry, which are so characteristic of Crohn's disease. Skip lesions are uncommon in tuberculosis. Colonoscopy with multiple biopsies may be helpful, as caseating granulomata containing acid-fast organisms can be found in a few patients. Finally, laparoscopy may show serosal or peritoneal tubercules.

Obviously, active tuberculosis elsewhere (such as lungs and kidney) makes the diagnosis of the intestinal disease, and hence the management, straightforward.

Perforation, acute dilatation, and massive haemorrhage may all occur in small-intestinal Crohn's disease but they are rare, being much less common than in Crohn's disease of the colon. Obstruction due to fibrous strictures is the most common local complication and the strictures may be multiple. Fistulae to neighbouring loops of small or large bowel or to other organs such as the bladder or vagina occur in 10 to 17 per cent of patients.

Gallstones are more frequently found in patients with ileal Crohn's disease, particularly if an ileal segment has been excised, than in a healthy population. The mechanism is the interruption of the enterohepatic circulation of the bile salts, with eventual depletion of bile salts.

Renal stones are also more commonly present in patients with Crohn's disease. They are usually oxalate stones, secondary to hyperoxaluria. This occurs when there is steatorrhoea since the fat binds intraluminal calcium. Thus the normal precipitation of calcium oxalate does not occur, leaving free oxalate to be absorbed from the colon and excreted through the kidneys. Another renal complication is the involvement of the right ureter in the inflammatory process occurring in the terminal ileum and caecum. This may give rise to a recurrent pyelonephritis or, more commonly, a right hydronephrosis.

Amyloidosis is a rare but well-recognized complication of long-standing Crohn's disease. When it occurs, it is often present throughout the intestinal mucosa and may also affect the kidney, leading to nephrotic syndrome and renal failure. If renal amyloid occurs, it is important to get the Crohn's disease into remission (that usually means resection) as the amyloid has been documented to regress subsequently.

Adenocarcinoma complicating small-intestinal Crohn's disease is very rare but is well recognized. A review of the 58 reported cases up to 1982 showed that the majority occurred in the ileum; 18 occurred in a bypassed loop. Most patients died within a year of diagnosis, presumably indicating the difficulty of making the diagnosis in a patient who already has Crohn's disease and abnormal radiological appearances of the small intestine.

Small-intestinal Crohn's disease, in contrast to colonic disease, is rarely associated with erythema nodosum, uveitis, or an acute arthropathy, which occur in no more than 1 per cent of patients with ileal disease. If these extraintestinal manifestations represent deposition of immune complexes, and the evidence for this is only suggestive, then it is possible that the greater antigenic load within the colonic lumen may explain their association with colonic disease. These manifestations occur when the intestinal disease is active.

Pyoderma gangrenosum may also complicate active Crohn's disease but, again, it is very rare when there is no colonic involvement.

Sacroiliitis occurs radiologically in 10 to 12 per cent of patients, although only a small proportion will complain of low back pain. Unless the patient has the HLA-B27 haplotype, sacroiliitis does not develop into a full-blown ankylosing spondylitis. This last condition occurs in about 1 per cent of patients with Crohn's disease who, almost exclusively, have the HLA-B27 haplotype. It is more commonly seen in patients with colonic involvement than in those with only ileal disease. Ankylosing spondylitis may present long before intestinal disease is suspected and is characterized by progressive stiffness of the back, leading ultimately to a rigid spine. The natural history of the back condition is independent of the Crohn's disease and, therefore, the presence of disabling back symptoms is not an indication for extensive surgical resection of the Crohn's disease.

Fatty liver may occur in any patient who is severely ill from Crohn's disease and is reversible once remission is achieved. During these severe attacks there may be transient rises in serum concentrations of transaminases or alkaline phosphatase, but they have no clinical significance.

Isolated granulomata may occur within the liver. Primary sclerosing cholangitis seems to be much less common in Crohn's disease than in ulcerative colitis, and is usually diagnosed on the basis of a persistently raised serum alkaline phosphatase. Liver biopsy may reveal the classical periductular fibrosis with chronic portal inflammation, but the appearances range from a chronic active hepatitis picture with portal inflammation and piecemeal necrosis, to gross fibrosis with obliteration of the bile ducts. The definitive diagnostic test is endoscopic retrograde cholangiopancreatography (ERCP), which will show the distribution of the disease—intrahepatic, extrahepatic, or a combination. Occasionally, there is a predominant stricture in the extrahepatic bile duct which can be worth dilating or stenting if the patient has an obstructive jaundice. Cholangiocarcinoma, a well-recognized complication of sclerosing cholangitis in patients with ulcerative colitis, is very rare in Crohn's disease and very few properly documented cases have been described. Whether this relates just to the low incidence of sclerosing cholangitis in Crohn's disease or whether other mechanisms are operating is unclear.

Since patients with Crohn's disease may have symptoms relating to active inflammation, bacterial overgrowth, obstruction, or as a result of previous surgery, assessment of the activity of the disease is difficult and a clinical assessment by itself can be misleading. Laboratory data frequently help and should include the haemoglobin, platelet count, albumin, and erythrocyte sedimentation rate (ESR). When available, serum concentrations of the acute-phase proteins, C-reactive protein and orosomucoid are particularly good indicators of active inflammation.

A large number of activity indices have been devised in order to standardize activity assessment. The Crohn's Disease Activity Index is the most widely used, as well as the Harvey–Bradshaw simplification of the Crohn's Disease Activity Index. However, the calculation of activity indices is subject to wide variation and their use is mainly for clinical trials rather than in routine clinical practice.

Labelled white-cell scans are often useful in patients when there is still doubt about whether symptoms are primarily inflammatory or obstructive. The distinction is obviously important for management decisions.

Since Crohn's disease cannot be cured, the role of the clinician is to control the inflammation, to correct nutritional deficiencies, and to ameliorate symptoms. These aims will frequently involve surgery and, indeed, 70 to 75 per cent of patients will require at least one operation during their lifetime. Thus, management of these patients requires close co-operation between physicians and surgeons. In general, medical treatment should be dictated by symptoms. Large doses of corticosteroids or immunosuppression are not indicated in asymptomatic patients who have active disease shown only by laboratory data.

Patients with severe disease are often malnourished and will require both nutritional replacement and support during the acute illness. Nutritional support may be parenteral or enteral, the latter being administered through a fine-bore nasogastric tube.

Parenteral nutrition by itself has no primary effect on the disease process, but several trials have now shown that elemental diets are as effective as corticosteroids in obtaining remission. However, the unpalatability of elemental diets makes for poor patient compliance and there is some evidence that relapses occur more quickly after dietary therapy than following corticosteroids. In the long term, patients should be advised to eat a normal well-balanced diet, avoiding only those foods which they know upset them.

A few patients with extensive small-bowel disease may have hypolactasia and will benefit from a lactose-free diet. The use of elimination diets is still unproven. However, a low-residue diet is advisable for patients with stricturing disease.

For most patients who have intermittent relapses of distal ileal disease, vitamin supplements are unnecessary, provided a serum vitamin B&sub1;&sub2; concentration is measured regularly (an annual measurement should be sufficient unless the value is progressively falling). However, patients with more extensive disease may benefit from folic acid, iron, B vitamins, and calcium. Magnesium and zinc deficiencies may also occur, and serum concentrations should be measured in patients with extensive disease that is chronically active. Oral iron is frequently poorly tolerated by these patients, and a total dose of iron given by slow intravenous infusion is often indicated. There is a potential danger of anaphylaxis but this occurs very rarely. Nevertheless, it is a wise precaution to have adrenaline and hydrocortisone readily available and the patient closely monitored while the infusion is in progress.

Patients with mild symptoms and with only minor changes in their inflammatory markers (such as platelet count, C-reactive protein, ESR) can be treated initially with sulphasalazine (1 g twice a day) or mesalazine for 4 to 6 weeks. These drugs have been shown to be effective in clinical trials, especially for colonic disease. Metronidazole (400 mg three times a day) may also be beneficial for mild disease. Alternatively, patients can be given oral prednisolone (20 mg daily) for 4 to 6 weeks.

For more active disease, or when initial treatment with sulphasalazine or mesalazine has not controlled symptoms, larger doses of prednisolone should be used. This usually implies starting with an oral dose of 40 to 60 mg daily, tapering to 20 mg over a period of 2 to 3 weeks. Patients are then treated for a further 4 weeks before the prednisolone is tailed off completely. If the disease is going to respond to cortocosteroids, that is usually complete by 6 weeks of therapy. In formal clinical trials, there has been no advantage in combining corticosteroid therapy with sulphasalazine.

Severe Crohn's disease is defined as a patient with symptoms of active disease (diarrhoea, abdominal pain, anorexia, weight loss) who has evidence of systemic disturbance (tachycardia, fever, anaemia). An abdominal mass may be present. These patients are best admitted to hospital, and they respond well to intravenous corticosteroids (for example, hydrocortisone, 100 mg every 6 h; or prednisolone, 20 mg every 8 h), intravenous fluids, and electrolyte and blood replacement. Intravenous metronidazole (500 mg every 8 h) is also often given, although there is no firm evidence to support its use. The majority of patients settle rapidly on this regimen and can be started on a light diet after a few days.

For patients with severe ileal disease complicated by stricturing or fistulae, intravenous corticosteroids and metronidazole should be given initially to allow much of the inflammation to settle before proceeding to surgery.

Parenteral nutrition, while not indicated for all patients requiring intravenous corticosteroid therapy, should be considered for each patient, especially if complications such as fistulae are present and if surgery seems a likely outcome.

Some patients go into a satisfactory remission and become symptomatic when their prednisolone dosage falls below 15 or 10 mg daily. Yet other patients rapidly relapse once prednisolone is withdrawn completely. These patients can be considered to have chronic active disease and it is this group that may benefit from immunosuppressive therapy.

In Europe, azathioprine (2 mg/kg) is usually used, but in the United States of America 6-mercaptopurine is more commonly used (azathioprine is metabolized to 6-mercaptopurine in the liver). Some patients respond well to these drugs and it is then possible to withdraw the corticosteroids. The exact mode of action of these drugs is not known but their beneficial effect, if it occurs, is seen over a period of several weeks. If a patient responds well, the problem then is to know how long to continue treatment. There are no data on this point but most clinicians would treat for 1 to 2 years before stopping the immunosuppressives.

Side-effects are fairly uncommon, with nausea and diarrhoea being the most frequent. However, regular blood counts (every 1–2 months) should be done, as marrow suppression occasionally occurs. Minor elevations of transaminase and an acute pancreatitis are other potential adverse reactions.

Recently, cyclosporin A has been used for chronic active disease at a dose of 5 mg/kg. Its role in therapy is still being evaluated in clinical trials. It is not recommended for general use at the present time since adverse reactions, in particular nephrotoxicity, are common. Methotrexate is also being evaluated but, so far, experience is limited to anecdotal reports only.

Diarrhoea in patients with Crohn's disease is best treated by treating the cause (active disease, bacterial overgrowth, etc). Patients with a bile salt-induced diarrhoea following ileal resection may benefit from cholestyramine but this must be used sparingly if more than 100 cm of terminal ileum has been resected as it will exacerbate fat excretion. However, a few patients will continue to complain of diarrhoea when all treatable causes have been identified and corrected. Antidiarrhoeal drugs (loperamide, codeine phosphate, diphenoxylate hydrochloride) may then have a role.

In contrast to ulcerative colitis, there is very little evidence that sulphasalazine is useful in maintaining remission, although a recent multicentre trial from Canada provides some support for the effectiveness of mesalazine. Several trials have tested the hypothesis that sulphasalazine or mesalazine might reduce the relapse rate following surgical resection. However, the results have been variable, with only some trials showing benefit. Therefore, until unequivocal evidence emerges for or against maintenance therapy with 5-aminosalicylic acid drugs, the decision whether to use such treatment remains an individual one. Certainly, there is no general case to the made for maintenance therapy with corticosteroids, although in the American National Cooperative Study, patients who went into remission on prednisolone appeared to stay in remission more frequently if low-dose corticosteroids were continued, compared to those who receive placebo maintenance therapy.

Dietary studies have shown that the subsequent course of the disease is not influenced by the amount of dietary fibre or sugar consumed.

Crohn's disease is a disease of high morbidity and low mortality. Nevertheless, despite recurrent relapses or, indeed, surgical operations, the majority of patients lead full and active lives if there has been good nutritional support, regular follow-up, and judicious use of medical and surgical treatment.

Following ileal resection, at least 45 per cent of patients will have a recurrence within 5 to 10 years, which tends to be less than the recurrence rate following ileocolonic resection. Of those that do relapse, approximately half will require further surgery. However, the risk of requiring further resections is probably not increased by previous resections, although this point is debated.

Recently, endoscopic examination of the ‘neoterminal ileum’ has shown recurrent inflammation in 65 to 80 per cent of patients during the 6 months following ileocolonic resection. Endoscopic appearances range from diffuse inflammation to aphthoid ulceration to confluent ulceration. However, it is not yet entirely clear whether the severity of these endoscopic recurrences predicts the subsequent clinical course.

Survival curves drawn by actuarial methods have shown that mortality tends to increase with the length of history. For the first 15 years of the disease, mortality is no different from a matched control population, but then tends to increase above that expected.

Female patients with Crohn's disease are as fertile, in general, as a control population. However, patients should avoid conceiving when the disease is active as there tends to be a high incidence of miscarriage. Apart from this, pregnancy seems to have little effect on the Crohn's disease and therefore patients should not be discouraged from having children if they wish. Should a relapse occur during pregnancy, it should be treated as vigorously as if the patient was not pregnant. Corticosteroids and sulphasalazine should be used as indicated, and should in no way be withheld. Patients conceiving while on treatment should be reassured that the developing fetus is not at risk, and the drugs should only be withdrawn as dictated by the activity of the disease. Some patients conceive while on azathioprine. Once again, there is no hard evidence that this will harm the fetus, although it is probably wise to stop the drug. However, if the Crohn's disease has been difficult to control and has only recently gone into remission with the introduction of azathioprine, the drug should be continued. A recent audit on the outcome of pregnancies in women with Crohn's disease who were receiving azathioprine has not shown an increased incidence of congenital abnormalities.

The disease can occur rarely within the first few months of life, but the frequency then increases after the age of 2 years. Epidemiological studies in the United Kingdom suggest that the incidence of Crohn's disease in children is increasing at a time when it has levelled out, or even fallen, in the adult population. The anatomical distribution of the disease is similar to that seen in adults. Symptomatically, the presentation may also be similar to that in adults, but some children will present with growth failure as the only manifestation. There is frequently a considerable delay in diagnosis, either because of the absence of gastrointestinal symptoms or because the children are labelled as having ‘toddler's diarrhoea’, ‘growing pains’, milk allergy, or even a functional syndrome. Once the diagnosis is suspected, it should be confirmed with full gastrointestinal assessment as described for adults.

Treatment should follow the same guidelines as those outlined for adults, with appropriate modification in the dosage of the drugs. Nutritional support is perhaps even more important than in adults, and a dietitian should document the child's dietary intake at regular intervals. Growth charts should be in the hospital notes, and recordings of height and weight should be made at each visit. Dietary supplements may be necessary to maintain an intake of at least 2000 kcal with 15 g nitrogen daily.

Allan RN, Keighley MRB, Alexander-Williams, J, Hawkins C, eds. Inflammatory bowel diseases. 2nd edn. Edinburgh: Churchill Livingstone, 1990.

Alstead EM, Ritchie, JK, Lennard-Jones JE, Farthing MJG, Clark ML. Safety of azathioprine in pregnancy in inflammatory bowel disease. Gastroenterology 1990; 99: 443–6.

Jewell DP, ed. Pathogenesis of inflammatory bowel disease. Eur J Gastroenterol Hepatol 1990; 2: 235–65.

Kirsner JB, Shorter RG, eds. Inflammatory bowel disease. 3rd edn. Philadelphia: Lea and Febiger, 1988.

Lowes JR, Jewell DP. The immunology of inflammatory bowel disease. Springer Semin Immunopath 1990; 12: 251–68.