Sustained clinical response to infliximab in refractory Cronkhite-Canada syndrome

Background

Cronkhite-Canada syndrome (CCS) is a rare, sporadic, non-hereditary gastrointestinal polyposis syndrome first reported in 1955 by Leonard Cronkhite Jr and Wilma Canada.1 It is characterised by diffuse hamartomatous gastrointestinal polyposis which results in malabsorption and protein-losing enteropathy. Clinical presentation is varied, and associated ectodermal manifestations include dysgeusia, alopecia, cutaneous hyperpigmentation and onychodystrophy—possibly relating to profound malnutrition. Characteristic pathology of CCS polyps consists of cystic gland dilation with stromal oedema and inflammation of the lamina propria. Diagnosis depends on clinicopathological correlation of endoscopic, histological and cutaneous features.2 3

Literature remains scarce regarding this rare condition, with approximately 500 cases described in small case series and case reports. About 75% of cases have been diagnosed in Japan.3 4 Little is known about the aetiology or pathogenesis. It is postulated that CCS is autoimmune in nature based on studies showing increased IgG₄ mononuclear cell staining in CCS polyps.5 There is also some preliminary work from Asia suggesting a relationship with Helicobacter pylori infection with case reports describing disease remission following eradication.6

Case presentation

A previously well 59-year-old Caucasian man presented on 11 November 2013 with a 10-month history of profound dysgeusia, nausea, anorexia and 18 kg weight loss. He was particularly bothered by a highly unpleasant metallic taste, and described food tasting like tin foil. He also reported intermittent epigastric pain and watery, non-bloody diarrhoea up to six times per day. The only pertinent family history was that of colorectal cancer in his father and maternal grandfather in their 60s. He specifically denied a family history of polyposis syndromes or other malignancies. Examination findings included cachexia and dystrophic nail changes but without alopecia or characteristic mucocutaneous hyperpigmentation (figure 1). There were no physical findings suggestive of micronutrient deficiencies, chronic liver disease or inflammatory bowel disease (IBD). There was no organomegaly on thorough abdominal examination and no ascites.

Investigations

Blood tests including electrolytes, renal function, full blood count, autoimmune screen, protein electrophoresis, immunoglobulins and IgG₄ levels were all normal. Iron profile was normal on presentation but B₁₂ was low at 97. CT with contrast of abdomen and pelvis was unremarkable other than demonstrating a grossly thickened duodenal wall.

Gastroscopy revealed significantly thickened, erythematous folds in the gastric body with large regenerative polypoid nodules in the antrum (figure 2). Biopsies including a polypectomy sample showed marked inflammatory regenerative changes with hyperplastic foveolar glands which were cystically dilated and tortuous (figure 3). Rapid urease testing and histology were negative for H. pylori.

Colonoscopy showed markedly thickened, oedematous and erythematous folds throughout the entire colon without ulceration (figure 4). There were hyperplastic polyps in the right colon and rectum, however assessment was hindered by the degree of background mucosal abnormality. Histology revealed widespread acute and chronic mucosal inflammation with oedema and widespread cystically dilated glands in both flat and polypoid mucosa (figure 5). Two foci of low-grade dysplasia were noted on random biopsies (figure 6). A histological diagnosis of CCS was suspected by our local pathologists, and all histology kindly reviewed by pathologists across international sites. The clinical, endoscopic and histological findings were felt to be most in keeping with CCS.

Differential diagnosis

In addition to CCS, hamartomatous polyposis syndromes include Peutz-Jeghers syndrome, Cowden syndrome, juvenile polyposis syndrome and hereditary mixed polyposis syndrome. CCS is an acquired condition, whereas the other syndromes are inherited in an autosomal-dominant pattern. For this patient, his age of disease onset, absence of family history, characteristic clinical features of dysgeusia and dystrophic nail changes, absence of mucocutaneous lesions and absence of extraintestinal malignancies all point towards a clinical diagnosis of CCS. This is supported by his characteristic histological features of marked submucosal oedema, widespread cystic glandular dilatation and hyperplasia of foveolar epithelium.

Treatment

The patient was commenced on 1 week of high-dose intravenous hydrocortisone which was weaned to oral prednisone as bridging therapy to azathioprine. He was concurrently treated with regular antiemetics, proton pump inhibitors, H1 and H2 antagonists, supplements including zinc, iron and multivitamins, and established on nasogastric enteral feeding. He also received 1 week of oral ciprofloxacin for possible small intestinal bacterial overgrowth without clinical response. Despite the above treatments, severe dysgeusia, abdominal pain and diarrhoea persisted. His nasogastric tube was converted to a nasojejunal tube but he remained intolerant of enteral feeding with further 8 kg weight loss.

Following failure of enteral nutrition, total parental nutrition (TPN) was started and gut rest established. His nausea and diarrhoea significantly improved however he was unable to regain weight, with further complications of multiple septic episodes from peripherally inserted central catheter infections. Through their research, the patient’s family discovered a blog describing the success of infliximab in a similar case. Based on this report, expert opinion was sought and infliximab was trialled. Azathioprine was continued at therapeutic doses for IBD treatment with regular 6TGN monitoring.

The patient received standard infliximab induction (5 mg/kg at 0, 2, 6 weeks) then eight weekly maintenance infusions. Within 3 weeks of his first infliximab dose, there was markedly increased tolerance of oral intake. His initial treatment with infliximab was complicated by recurrent septicaemia. These included three admissions with central line associated bacteraemia (Staphylococcus aureus, Serratia marcescens and Pseudomonas aeruginosa), as well as one admission with Enterobacter urosepsis. His TPN was stopped following 4 months of infliximab treatment, during his last hospital admission with tunnelled line sepsis. By that stage he was able to maintain weight by oral diet alone.

Outcome and follow-up

After 12 months of infliximab treatment, he was in complete clinical remission and returned to his baseline weight. All dietary supplements were stopped. Alongside this significant clinical response there were also dramatic endoscopic improvements. Endoscopies were performed at 6 months, 18 months, 3.5 years and 5 years following infliximab therapy. Gastroscopy at 6 months already showed a significant regression of gastritis and duodenitis, however multiple large antral polyps remained, with a maximum diameter of 20 mm. Colonoscopy at 6 months showed normalisation of colonic mucosa but multiple hyperplastic polyps remained. Subsequent endoscopies showed significant polyposis regression. Gastroscopy at 18 months showed the largest polyp had reduced to 10 mm, whereas by 3.5 years diminutive sessile gastric polyps remained with a single 6 mm pedunculated duodenal juvenile-type polyp which was fully resected. Gastroscopy at 5 years showed only diminutive polyps in the gastric body, with some features suggestive of hamartomatous polyps on biopsy (figure 7). Colonoscopies at 3.5 years and 5 years showed complete resolution of previous polyps, normal colonic mucosa and normal terminal ileum on deep intubation (figure 8). Random colonic biopsies were histologically normal and demonstrated deep remission.

The patient has remained in clinical remission for 6 years while continuing combination treatment with infliximab and azathioprine. 6TGN levels remain therapeutic at 413 pmol/8×10⁸ red blood cell (red cell count 4.7×10¹²/L), however latest infliximab trough levels are subtherapeutic (0.7 mg/L) and infliximab antibodies have been detected.

There have been regular discussions with the patient since achieving clinical and endoscopic remission about the risk–benefit balance of reducing his immunosuppression regime. At present, despite the low infliximab levels and presence of infliximab antibodies, treatment has been continued given the associated morbidity to be expected with a possible disease relapse. The patient also feels strongly about continuing treatment given how critically unwell he had been prior to infliximab therapy.