Colonic endocrine cells in patients with familial amyloidotic polyneuropathy


Dr M. El-Salhy, Department of Medicine, Umeå University Hospital, S-901 85 Umeå, Sweden (fax: +46 90 143986; e-mail:


Abstract. Anan I, El-Salhy M, Ando Y, Nyhlin N, Terazaki H, Sakashita N, Suhr O (University Hospital, Umeå, Sweden, and University Hospital, Kumamoto, Japan). Colonic endocrine cells in patients with familial amyloidotic polyneuropathy. J Intern Med 1999; 245: 469–473.

Objective.  To establish whether the endocrine cell number is affected in the colon in Japanese FAP patients.

Setting.  Department of Medicine, Umeå University Hospital and Department of Internal Medicine and Pathology, University Hospital, Kumamoto, Japan.

Subjects.  Autopsy colon tissue specimens from 11 FAP patients and nine controls as well as 12 control biopsy specimens were included in the study.

Measurements.  Endocrine cells in the colon were detected by immunohistochemistry and quantified by computerized image analysis.

Results.  The autopsy material showed a slight autolysis. Neither enteroglucagon nor pancreatic polypeptide positive cells could be detected in the autopsy material, but were present in biopsy material. There was no statistical difference between autopsy and biopsy specimens regarding the number of peptide YY (PYY), somatostatin and serotonin cells. No significant differences were noted in PYY, somatostatin and serotonin immunoreactive cells in FAP patients compared to autopsy controls, though PYY cells tended to be decreased and serotonin and somatostatin cells tended to be increased in FAP patients.

Conclusion.  The difference between the Swedish and Japanese patients in the endocrine cell content points to the possibility of involvement of other factors than the endocrine cell depletion of the colon might be involved in the pathogenesis of gastrointestinal dysfunction in FAP. The tendency of PYY to decrease in Japanese FAP might contribute to the development of diarrhoea in these patients.


Familial amyloidosis with polyneuropathy type I (Portuguese or Andrade type) is an autosomal dominant form of systemic amyloidosis, which was first described in 1952 by Andrade [1]. This type of amyloidosis is geographically clustered in Portugal [1], northern Sweden [2] and Japan [3], but sporadic cases have been reported in other countries such as Ireland, USA and China. Familial amyloidosis with polyneuropathy (FAP) is caused by mutated transthyretin (TTR), in which valine is replaced by methionine in position 30 [4]. The disease is characterized by a progressive polyneuropathy that affects both the peripheral and the autonomic nervous systems [5]. In Swedish FAP patients, the initial clinical manifestation is a sensory motor polyneuropathy starting in the lower extremities and later affecting the upper ones. Symptoms from the autonomic nervous system include orthostatic hypotension, anhidrosis, sexual impotence, heart dysrhythmia and gastrointestinal disturbances [2]. Initial gastrointestinal symptoms are severe constipation, often accompanied by nausea and vomiting. Gradually, the constipation is relieved by periods of diarrhoea which later become continuous [6]. The cause of death is usually intercurrent infections and malnutrition.

Although the mutation of TTR in Japanese and Swedish FAP patients is identical, they disclose different clinical manifestations of the disease. Whilst the mean age of onset in Swedish patients is the middle fifties, that of Japanese patients is middle thirties [7]. In the majority of Japanese FAP patients, autonomic nerves are impaired before sensory and motor nerves [8]. Most Japanese patients exhibit gastrointestinal symptoms such as diarrhoea, constipation, alternating diarrhoea and constipation, nausea and vomiting during the course of the disease. Amongst these symptoms, diarrhoea is more frequently found early in Japanese than Swedish patients [7].

The neuroendocrine system in the digestive tract plays an important role in regulating motility, secretion, blood flow and ion transport [9]. Recently a reduction of several endocrine cell types has been found in the duodenum and rectum in Swedish FAP patients [1011] and it has been suggested that these abnormalities are a contributing factor in the development of gastrointestinal disturbances. The present study was undertaken to investigate if similar reduction of the endocrine cells is present in the colon of Japanese end stage FAP patients.

Materials and methods

Patients and controls

Autopsy material from the sigmoid colon of 11 FAP patients (six females and five males with a mean of age 46.2 years; range 34–63 years) was investigated. In all cases, the diagnosis had been verified by the finding of amyloid deposits in biopsy specimens from skin or intestinal mucosa. Furthermore, the diagnosis was supported by DNA analysis for the transthyretin mutation of valine to methionine at position 30. All patients had gastrointestinal symptoms (details concerning clinical data are summarized in Table 1). As controls, autopsy material from the sigmoid colon was obtained from nine subjects (five females and four males, mean age 59.9; range 40–85 years). Five of these subjects died of colon cancer, one of acute myocardial infarction, one of hepatoma, one of lung cancer and one of adult T-cell leukaemia. To evaluate the impact of autolysis on the autopsy material, macroscopically and histologically normal tissues from the colon of 12 patients (eight females and four males; mean age 66 years; range 34–86 years) were included. Six had undergone colon resection because of polyps, chronic diverticulitis, prolapsis and volvulus. The tissue samples taken from these control patients were from the sigmoid colon, except for one from the ascending colon. The remaining six patients had undergone routine standard colonoscopy examination because of rectal bleeding, where haemorrhoids were identified as the source of bleeding. Biopsy specimens from these patients were taken from the mucosa of sigmoid colon.

Table 1.  Please [INSERT TABLE LEGEND] Thumbnail image of

Histopathological and immunohistochemical examination

Tissue specimens were fixed overnight in 4% buffered formaldehyde, embedded in paraffin and cut at 5 µm. Sections from both patients and controls were stained with haematoxylin-eosin, van-Gieson stain and alkaline Congo red stain. Detection of amyloid deposits was performed in polarized light.

The sections were immunostained by the avidin-biotin complex (ABC) method (DAKO A/S, Glostrup, Denmark) as described earlier in detail [12]. The primary antisera used were raised in rabbit against: peptide YY (PYY) (dilution 1 : 2000, code no. R 841303, Eurodiagnostica, Malmö, Sweden) pancreatic polypeptide (PP) (dilution 1 : 500, code no. A 619, Dakopatts), enteroglucagon (dilution 1 : 500, code no. R 781101, Eurodiagnostica), somatostatin (dilution 1 : 6000, code no. A 566, Dakopatts), and serotonin (dilution 1 : 400, code no. M 0758, Dakopatts). Specificity controls were the same as described previously [12].

Computerized image analysis

Morphometric measurements were performed by Quantimet 500 MC image Processing and Analysis System (Leica, Cambridge, England) linked to an Olympus microscope, type BX50. The software used in this system was the Leica Windows-based image analysis program QWIN (version 1.02) and interactive programming system QUIPS. The immunostained sections of colon from controls and patients were coded and mixed, and image analysis was performed without knowledge of the identity of the sections. Quantification was done by X20 objective in 20 randomly chosen fields as described previously [13] using an automatic standard sequence analysis operation.

Statistical analysis

Comparison between groups was carried out with the Mann–Whitney test. P-values below 0.05 were considered as statistically significant.


Histopathological examination

Examination of haematoxylin-and Van-Gieson-stained sections both from controls and patients showed minor autolytic changes: epithelial cells showed pyknotic nuclei and vacuolated cytoplasm, occasionally completely autolysed cells were found, especially from one control subject, where tissue specimens were taken 6 h after death. Amyloid deposits were found in all patients; in the submucosa, especially in the wall of blood vessels.


The endocrine cells detected were localized in the upper middle part of the crypts, and were of varying shape, from oval to triangular. PYY- ( Fig. 1), somatostatin- and serotonin-immunoreactive cells were observed in the colonic autopsy specimens from both controls and FAP patients. Enteroglucagon-and PP-immunoreactive cells were neither detected in autopsy specimens from controls nor from FAP patients. In the non-autopsy controls both enteroglucagon and PP-immunoreactive cells were present.

Figure 1.

Peptide YY (PYY)-immunoreactive cells in the colon of an autopsy control (a) and a patient with familial amyloidosis (FAP) (b). ×180.

Computerized image analysis

The results of the morphometric measurements of the endocrine cells are shown in Fig. 2. There was no statistical difference between biopsy and autopsy specimens as regards the number of PYY-, somatostatin- and serotonin-immunoreactive cells (P = 0.4, 0.6 and 0.7, respectively). Whereas the number of PYY showed a tendency to decrease, somatostatin- and serotonin-immunoreactive cells tended to increase in FAP patients compared to autopsy controls. These changes were, however, not statistically significant (P = 0.2, 0.5 and 0.7, respectively).

Figure 2.

The number of different endocrine cell types in the colon. Controls I represent autopsy samples and controls II represent biopsy samples from resected colon material. (●) is median, the boxes represent 25 and 75 percentiles.


The present investigation was performed on autopsy specimens from the colons of Japanese patients with FAP. Despite the relatively short time between death and specimen fixation, the tissue specimens showed minor autolytic changes, and neither PP nor enteroglucagon cells could be detected in autopsy material, though they were present in biopsy material. However, the number of detectable endocrine cell types in autopsy specimens: PYY, somatostatin and serotonin did not differ significantly between autopsy and non-autopsy material.

It would have been desirable to study fresh biopsy colon specimens from FAP patients. Unfortunately, such specimens are difficult to obtain since the condition of these patients seldom calls for colon resections or colonoscopy.

The PYY immunoreactive cells in the colons of Japanese patients showed a tendency to decrease in number. This finding corresponds to that of Swedish FAP patients in rectal immunoreactive cells [11]. However, the number of serotonin and somatostatin cells was decreased in the rectums of the Swedish FAP patients [11], whereas the number of these cells in the present study tended to be increased. However, the present investigation differs from the study on Swedish FAP patients in several aspects. The part of the large bowel investigated in the Swedish patients, i.e. the rectum, contains a larger number of endocrine cells than the colon [14]. Moreover, the Swedish patients had a short duration of the disease, and constipation was the major gastrointestinal complaint. In contrast, patients in this study had a long duration of the disease and diarrhoea was the major gastrointestinal complaint. These differences may contribute to divergent findings between the two studies. It has been suggested, however, that environmental and/or genetic factor(s) have impact on the clinical pattern of FAP [15]. Therefore, the difference in colonic endocrine cells between Swedish and Japanese FAP patients may be related to environmental or genetic differences.

Peptide YY delays intestinal transit and is considered to be one of the mediators of the ileal brake [16]. Moreover, it has an antisecretory effect [16]. Serotonin has been found to stimulate motility in the small intestine and colon, as well as to accelerate transit through the small and large intestine [17]. Somatostatin inhibits ileal longitudinal muscle contraction, but stimulates intestinal migrating motor complexes. It also inhibits secretion of fluid and bicarbonate and absorption of calcium, glycerol, amino acids and triglycerides [18]. It is tempting to speculate that the tendency for PYY cells to increase in Japanese FAP patients might contribute to the development of diarrhoea in these patients.

The present study indicates that other factors than endocrine cell number in the colon might be involved in the pathogenesis of gastrointestinal dysfunction in FAP.


This study was supported by Grants from the Swedish Medical Association, Umeå Health District, the medical faculty of Umeå University, and FAMY.

Received 12 January 1998; accepted 15 September 1998.