Article Text

Primary antibody deficiency and Crohn's disease
  1. P Conlong,
  2. Wynne Rees,
  3. J L Shaffer,
  4. David Nicholson,
  5. Derek Jewell,
  6. Mansel Heaney,
  7. Aled Jones,
  8. Neil Snowden
  1. Hope Hospital, Stott Lane, Salford M6 8HD, UK
  1. Dr P Conlong, 14 Lidgate Grove, Didsbury, Manchester M20 6ST, UK


Five patients with primary antibody deficiency were investigated because of intermittent but persistent diarrhoea of several years duration despite immunoglobulin replacement therapy. We found no evidence of Giardia lambia or other intestinal pathogens to explain their gastrointestinal symptoms. All five had definite radiological evidence of small bowel Crohn's disease and three had histological specimens available with abnormalities consistent with Crohn's disease. One patient had a non-caseating granuloma in an oral ulcer. A second patient with stricturing disease in the small bowel had a mucosal inflammatory infiltrate with non-caseating granulomas. A third had transmural inflammation but no granulomas. All five patents were diagnosed as having Crohn's disease and have responded symptomatically to steroid therapy.

  • antibody deficiency
  • Crohn's disease

Statistics from

Some patients with primary antibody deficiency have an enteropathy. Infection is common and although immunoglobulin replacement therapy seems to have reduced the incidence of some bowel pathogens such as Giardia, Salmonella andCampylobacter, 10% have diarrhoea, abdominal pain or weight loss which defies explanation.1

There is a broad spectrum of histological abnormality in these patients. In the stomach there is a high frequency of gastric atrophy and a predisposition to gastric carcinoma,2-4whilst in the small bowel, lymphoid nodular hyperplasia is common.2 3 5 In those with common variable immunodeficiency, there are well-documented cases of jejunal villous atrophy which may or may not respond to gluten withdrawal; however, the diagnosis of coeliac disease has only been reliably proved in one patient after gluten withdrawal and challenge.6

There have been several case reports showing radiological appearances compatible with Crohn's disease in patients who have presented with diarrhoea and intestinal strictures, although the histology has generally not shown granulomas or transmural inflammation.7-11 Teahon and Webster,11 in their series of patients with primary antibody deficiency, proposed the existence of an enteropathy distinct from Crohn's disease. The radiology was consistent with Crohn's disease but the histology was apparently distinctive with the presence of microscopic colitis, increased intra-epithelial lymphocytes and an intact crypt architecture in the colon. Any inflammation in the small bowel was not transmural and there were no granulomas.

We investigated the cause of diarrhoea in five patients with primary antibody deficiency who were referred with refractory diarrhoea.


We investigated five patients who were referred with intermittent but persistent diarrhoea of several years duration from a cohort of 100 patients with primary antibody deficiency from the Department of Immunology at Hope Hospital in Manchester over a period of 8 years. There were four men (19–65 years) and one woman (54 years). One had X-linked agammaglobulinaemia (Bruton's disease), the others had common variable immunodeficiency. All five patients had been shown to have a primary immunological deficiency by their repeated failure to generate responses to common antigens such as pneumococcus, tetanus and pokeweed nitrogen. All were on regular immunoglobulin replacement therapy with Sandoglobulin (Sandoz Pharmaceuticals) at a dose of 0.3–0.4 g/kg/month with regular assessment of their immunoglobulin trough levels.

Two sets of three fresh stool specimens from each patient were examined for evidence of parasitic involvement or other pathogens. Each patient had a gastroscopy with duodenal biopsy, full colonoscopy and a small bowel enema. The specimens obtained were examined histologically for parasites.

Patients were commenced on steroid therapy and their response assessed according to improvements in their symptoms. They were then followed up over a 5-year period.


The main findings are shown the table. No pathogens were identified from several stool examinations specifically looking forSalmonella, Shigella, Campylobacter, ova and cysts. All of the patients had radiological evidence of small bowel abnormalities compatible with Crohn's disease. Patient 1 had nodular thickening throughout the jejunum and ileum (figures 1 and 2). Patient 2 showed stricturing disease in the duodenum and small bowel (figure3). Patient 3 had a stenotic lesion in the terminal ileum with skip lesions and ulceration. Patient 4 had rosethorn ulceration of the terminal ileum. Patient 5 also had stricturing disease in the terminal ileum (figure4).

Figure 1

Film on barium follow-through study (patient 1). There is mild thickening and nodularity of the jejunum. In addition in this case more diseased loops of bowel are seen within the pelvis where the nodularity is more marked (open white arrow) and there is associated luminal narrowing with bowel loop separation (black arrow)

Figure 2

Film from small bowel enema (patient 2). Note the small bowel enema tip in third part of duodenum. Strictures are evident: the first (white arrow) is at the junction of the second and third part of the duodenum. A second stricture is seen in the region of the DJ flexure (curved open arrow). This is causing pre-stenotic dilatation of the third part of the duodenum which contains a filling defect/food residue. Contrast has, however, passed beyond this point into the proximal jejunum

Figure 3

Film from small bowel follow-through series (patient 2). There is mild, diffuse nodular thickening of the jejunum and proximal ileum. No structuring disease is identified. The appearances are non-specific but can be seen in diffuse Crohn's jejunal ileitis as well as many other conditions such as Giardiasis.

Figure 4

Cone view of right iliac fossa during small bowel enema (patient 5). The distal terminal ileum shows stricturing in several places (white arrows). The folds are thickened and there is nodularity with several adverse ulcers. These are most prominent between the two labelled strictures.

There was a wide spectrum of histological abnormality. Patient 1 had granulomatous ulceration in the mouth and duodenitis. Patient 3 had total villous atrophy on duodenal biopsy with increased intra-epithelial lymphocytes and crypt hyperplasia. Patient 5 had gastritis. Patients 2 and 5 needed resections for obstructive symptoms. One resected terminal ileal specimen revealed an inflammatory non-transmural lymphocytic infiltrate, fibrosis and two non-caseating granulomas. The other resected ileal specimen showed transmural lymphocytic inflammation without granulomas. In the colon of patients 3 and 4 the specimens demonstrated a mild non-specific inflammation restricted to the mucosa which was mainly lymphocytic with some neutrophils, infiltration, no crypt abscesses, and without intra-epithelial lymphocytes. There was no evidence of parasites in any of the specimens examined.

Patients 1 and 3 who had particularly severe symptoms were treated with budesonide but did not respond and were controlled on regular prednisolone. The other three patients responded to short courses of prednisolone following commencement of steroids, a clinical improvement was documented by a significant reduction in stool frequency.


The five patients described had Crohn's disease. We found no evidence to support the existence of a separate enteropathy associated with primary antibody deficiency. One might have expected that infection would explain these patients relapsing and remitting gastrointestinal symptoms. After all, the majority had evidence of infection in other body systems, but despite looking for pathogens in stools and histological specimens none was ever implicated. It could be argued that we did not look aggressively enough for micro-organisms.Microsporidia andCyclospora are difficult to detect on haematoxylin and eosin sections; electron micrographs would perhaps have been preferable. It is also difficult to attain a perfect plateau of immunoglobulin replacement, even with frequent monitoring of trough levels, and Sandoglobulin contains predominantly IgG with only small amounts of IgA, the principal bowel immunoglobulin.

All of our patients had radiological evidence of small bowel involvement varying from stricturing disease requiring surgical resection in two patients to nodular thickening in the jejunum and ileum or rosethorn ulceration. The histological abnormalities were very diverse. Two patients had non-caseating granulomas; in one, this was present in a lesion in the mouth and the other in a resected terminal ileal specimen. This latter patient, however, had a predominantly mucosal rather than transmural lymphocytic infiltrate. The other patient who needed a terminal ileal resection for stricturing disease had a mild transmural lymphocytic infiltrate without granulomas.

Total villous atrophy with increased intra-epithelial lymphocyte infiltration was found in one patient on duodenal biopsy but he showed no improvement in histological appearance with gluten withdrawal. Another had lymphocytic infiltration in the duodenum but no other abnormalities. In colonic specimens, two patients had mild inflammation restricted to the mucosa, with an intact crypt architecture and no increase in intra-epithelial lymphocytes.

We would argue that the pattern of histological and radiological findings demonstrate these patients to have Crohn's disease rather than belonging to a different grouping as originally proposed by Teahon and Webster.11

Patients 2 and 5 needed small bowel resections for stricturing disease in the early part of their clinical course; their subsequent gastrointestinal symptoms have responded well to short courses of prednisolone, as have those of patient 4. Two other patients (1 and 3) were tried on 9 mg of budesonide but did not respond and have better disease control on regular prednisolone therapy, with one of these also tolerating and responding to an elemental diet. Patient 1 with x-linked agammaglobulinaemia has recently developed a protein-losing enteropathy initially suggested by rising requirements for immunoglobulin therapy. Interestingly, this problem appears to have been controlled with steroid therapy.

The aetiology of Crohn's disease is still unknown. A variety of hypotheses exist but none have been proved. Immune function has been extensively tested in Crohn's patients. There have been studies suggesting mild defects in neutrophil functions, both in ability to phagocytose12 and in chemotaxis on ‘skin windows’13 (serum-filled perspex windows placed over abraded skin). Others, however, have shown normal neutrophil oxidative ability14 15 and normal chemotaxis in vitro. All in all, these probably add up to mild defects in phagocyte function. Lymphocyte function is thought to be normal in Crohn's disease; although there is a specific increase in IgG2 immunocytes.16 It is not known how many patients with Crohn's have primary antibody deficiency.

In conclusion, we report five patients with primary antibody deficiency who also have Crohn's disease defined radiologically and histologically, and whose symptoms are controlled on steroid therapy.


Table 1
View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.