|
A correction has been published: N Engl J Med 1993;329(21):1592.
|
| |||||||||||||||||||||||||||||||||
Background and Methods The pathophysiology of diarrhea in patients with the carcinoid syndrome is not understood. Possible causes include tumor production of neurohumoral substances, such as serotonin and substance P, which stimulate small-bowel and colonic motility, and intestinal abnormalities, such as lymphangiectasia and bacterial overgrowth. We undertook this study to determine whether carcinoid diarrhea is associated with abnormal motor function in the small intestine and colon. We measured the gastric, small-bowel, and colonic transit of radiolabeled solid residue and estimated the volume of the ascending colon in 16 patients with the carcinoid syndrome and diarrhea and 16 normal subjects. We also measured colonic tone and phasic pressure activity by intracolonic multilumen manometry and with an electronic barostat in seven patients and six normal subjects.
Results The patients with the carcinoid syndrome had elevated 24-hour urinary excretion of 5-hydroxyindoleacetic acid and elevated fasting plasma serotonin concentrations. Transit times in the small bowel and colon were two times (P<0.001) and six times (P = 0.001) faster in the patients than in the normal subjects. The volume of the ascending colon was approximately 50 percent smaller in the patients than in the normal subjects (P<0.001). The patients had normal fasting colonic tone; their mean postprandial colonic tone was markedly increased as compared with the values in the normal subjects (mean increase, 41 percent vs. 24 percent; P = 0.03).
Conclusions Patients with the carcinoid syndrome who have diarrhea have major alterations in gut motor function that affect both the small intestine and colon.
We hypothesized that diarrhea in patients with the carcinoid syndrome results primarily from abnormal intestinal motor function and that the capacity of the proximal colon is reduced as a result of increased colonic tone. We therefore measured the transit of radiolabeled solids through the digestive tract, the volume of the ascending colon, and colonic tone and motility in 16 patients with this syndrome.
Methods
Study Subjects
We studied gastrointestinal transit in 16 consecutive patients (11 men and 5 women; age range, 30 to 77 years; mean, 62) who fulfilled the following entry criteria: histologically proved metastatic carcinoid disease, associated diarrhea (>5 bowel movements per day), and no previous intestinal resection or treatment with octreotide. None had other clinical signs of niacin deficiency, such as skin rash or dementia, or signs of intestinal obstruction. Twelve of the patients underwent laparotomy for resection or tumor debulking after our studies were completed. Urinary excretion of 5-hydroxyindoleacetic acid was measured in all patients, and fasting plasma concentrations of serotonin were measured by high-performance liquid chromatography and amperometric detection17 in 13 patients; substance P was measured by radioimmunoassay18 in 14 patients. Seven days before the studies began, the patients discontinued all antidiarrheal medications, such as opioids and clonidine, and other agents that might affect intestinal transit, such as calcium-channel-blocking drugs. Sixteen normal subjects (6 men and 10 women; age range, 33 to 73 years; mean, 54) were recruited by public advertisement.
Seven of the 16 patients with carcinoid syndrome (3 men and 4 women; age range, 30 to 77 years; mean, 62) participated in a study of colonic motility performed one day after the transit study. Six additional normal subjects (four men and two women; age range, 22 to 45 years; mean, 31) were recruited by public advertisement for this motility study.
The studies were approved by the Mayo Institutional Review Board, and written informed consent was obtained from all subjects.
Measurement of Gastrointestinal Transit
Transit through the entire gastrointestinal tract was measured by methods developed in our laboratory15,19. Technetium-99m-labeled Amberlite 410 resin pellets (Sigma Chemical, St. Louis) were added to an egg breakfast meal (219 kcal) and used to measure gastric and small-bowel transit,19 and a methacrylate-coated, delayed-release capsule containing indium-111-labeled Amberlite IR-120P resin pellets (Sigma Chemical) was used to measure colonic transit. Total-body exposure to radiation was 5 x 10-6 Gy. Four and eight hours after the breakfast, a standardized lunch (536 kcal) and dinner (561 kcal) were eaten. Anterior and posterior gamma-camera images (Starcam, General Electric, Milwaukee) were obtained at regular intervals to measure 99mTc and 111In radioactivity in the gastric and colonic regions. Four colonic regions of interest -- the ascending, transverse, descending, and rectosigmoid regions -- were imaged on sequential scintiscans of the colon as previously described (Figure 1)15,19. The counts were corrected for radionuclide decay and for the detection of 111In counts in the 99mTc window. The beginning of colonic transit was defined as the time at which all 111In counts were located in the colon. The timing and 111In content of each stool were recorded.
|
Calculation of Ascending-Colon Volume by Scintigraphy
The scans in each subject that showed the highest 111In counts within the ascending colon were divided transversely into slices 12.8 mm high, and the transverse diameter of each was calculated by full-width half-maximum analysis, which is independent of the absolute number of counts within each slice23. The area of each segment was calculated by the formula piD2/4, where D2 is the square of the transverse diameter. The volume of each 12.8-mm segment was then estimated (area x 12.8 mm); the sum of all the segments was the calculated volume of the ascending colon. The results of estimates of the volume of the ascending colon with this method are almost identical to those obtained with single-photon-emission computed tomography, a three-dimensional method24.
Measurement of Colonic Motility
On the evening before the study, the subjects drank a solution of polyethylene glycol and electrolytes (OCL, Abbott Laboratories, Chicago) until their fecal effluent consisted of clear liquid. The next day, colonoscopy was performed after the intravenous administration of 2 to 4 mg of midazolam (Versed, Roche Laboratories, Nutley, N.J.). A soft-tipped, Teflon-coated guide wire (Microvasive, Hobbs Medical, Stafford Springs, Conn.) was inserted into the cecum, and the colonoscope was withdrawn. A combined manometer and barostat (described below) was advanced into the colon over the guide wire with the aid of fluoroscopy. In three normal subjects and three patients, the barostat balloon was placed in the transverse colon; in the other subjects it was positioned in the descending colon just beyond the splenic flexure.
The manometer-barostat, adapted from one used in recent studies,25 had side holes 2 cm proximal and 2, 7, and 12 cm distal to the barostat balloon for pneumohydraulic perfusion manometry. The balloon was a highly compliant polyethylene bag, 10 cm in length (Hefty Baggies, Mobil Chemical, Pittsford, N.Y.), tied at both ends to metal rings incorporated in the device. Colonic tone was assessed by noting the changes in the balloon volume in the presence of an intraballoon pressure ranging from 12 to 17 mm Hg that was selected at the start of the study and kept constant throughout the study25.
The measurements were begun at least four hours after the subjects were sedated, by which time all subjects were alert and positioned in a 30-degree head-up, supine position. The fluctuations in balloon volumes and intraluminal pressures were recorded for 30 minutes before the participants drank a 750-ml, 1000-kcal chocolate drink (53 percent fat) and for 90 minutes afterward.
The pressures and volumes in the barostat balloon, pressure fluctuations during the manometric phase, respiratory movement, and experimental interventions were all recorded on paper and on a computer (Microvax System and modified Vaxlab program, Digital Equipment, Boston) for later filtering of respiratory and motion artifacts and analysis.
For the barostatic estimation of colonic tone, a computer program was used to separate base-line volume from phasic volume events (Figure 2). Base-line balloon volumes were calculated by excluding phasic volume events that coincided with pressure activity recorded during the manometric phase26. The average of the volumes measured for 30 minutes before the meal represented the fasting tone. The volumes were averaged for each five-minute period postprandially; the mean change in volume for each five-minute period relative to the fasting volume was calculated. Volume changes were also expressed as a percentage of the fasting volume to account for differences among subjects in fasting volumes, thereby providing an estimate of the overall postprandial change in colonic tone. The lowest balloon volume persisting for 10 or more consecutive minutes postprandially provided an estimate of the maximal colonic tone and was also expressed as a percentage of the fasting volume.
|
Statistical Analysis
Data on gastric-transit and small-bowel-transit times, colonic volumes, and fasting and postprandial colonic tone were compared with unpaired two-tailed t-tests (Clinfo, National Institutes of Health, Bethesda, Md.); the overall colonic-transit time (the geometric center of the isotope in the colon at 24 hours) and colonic-motility indexes did not pass a test for normality of distribution and were compared by Wilcoxon's two-tailed unpaired rank-sum test (Clinfo). The results are given as the means ±SE.
Results
Characteristics of the Patients
The characteristics of the 16 patients with metastatic carcinoid disease and diarrhea are summarized in Table 1. One patient had a foregut carcinoid tumor; the remainder had tumors that originated in the midgut. Two patients had prominent mesenteric fibrosis; the small-bowel resection margins were unaffected by tumor, and neither patient had evidence of intestinal lymphangiectasia. Urinary excretion of 5-hydroxyindoleacetic acid was increased in all patients. Fasting plasma serotonin concentrations were elevated in the 13 patients in whom they were measured, and substance P concentrations were elevated in 5 of the 14 patients in whom they were measured.
|
There was no significant difference in gastric-emptying times between the patients and the normal subjects (Table 2). In contrast, the transit time in the small bowel, overall colonic-transit time at 4 and 24 hours, and emptying rate of the proximal colon were all significantly faster in the patients.
|
|
The ascending-colon volumes were approximately 50 percent smaller on average in the patients than in the normal subjects (mean [±SE], 81 ±7 vs. 174 ±15 ml; P<0.001). The volumes were still low when remeasured later in Patients 4 and 5 (Table 3).
Colonic Tone
Examples of the colonic barostatic tracings in the patients and normal subjects are shown in Figure 3, and the measurements of colonic tone are summarized in Table 4. The fasting volumes were similar in the patients and normal subjects. The postprandial volumes were lower than the fasting volumes in both groups, but the mean change in volume was greater in the patients than in the normal subjects (41 percent vs. 24 percent, P = 0.03). Similarly, the maximal change in volume after the meal was greater in the patients than in the normal subjects (61 percent vs. 39 percent, P = 0.01).
|
|
Data on the pressure activity during the colonic phase are also shown in Table 4. The fasting motor indexes were not significantly different in the two groups; postprandially, the motor indexes increased in both groups relative to the fasting values (P<0.05), but the increases were similar in the two groups. There were no high-amplitude propagated contractions in either group.
Discussion
Rambaud et al.28 proposed the concept of "diarrhee motrice" (or motor diarrhea) after observing the accelerated transit of barium from the pylorus to the cecum and of carmine from ingestion to excretion in stool in two patients with the carcinoid syndrome11. One patient had accelerated transit in both the small bowel and the colon; the second had rapid transit in the small bowel. In two other patients manometry of the small bowel and colon showed a hyperactive small intestine and inactive colon12,13. Our results support the hypothesis of Rambaud et al.: overall colonic transit was accelerated; the emptying rate of the proximal colon was about six times faster and small-bowel transit was about two times faster in the patients with the carcinoid syndrome than in the normal subjects. In view of the finite capacity of the proximal colon to store15 and reabsorb16 an increased fluid load, our results also suggest that rapid emptying of the proximal colon may contribute to the increased stool weight in patients with the carcinoid syndrome.
Rapid emptying of the proximal colon is associated with a smaller volume in the ascending colon. It is conceivable that this reduced capacity results from the exaggerated postprandial tonic response in the colons of patients with the carcinoid syndrome. The mechanisms underlying these abnormal motor functions are still unclear. Luminal factors, such as fat or its metabolites,29 are unlikely causes, since none of the patients had a history of steatorrhea, intestinal resection, or intestinal lymphangiectasia. The motor dysfunction is more likely to be humorally mediated in view of the elevated plasma serotonin or substance P concentrations in these and other patients30. A neurally mediated gastrocolonic response may be enhanced by peptides, amines, or hormones secreted by the tumor postprandially. Sleisenger et al.31 have demonstrated serotoninergic modulation of cholinergic responses in the canine gut. In future studies, these and other substances (such as motilin and cholecystokinin, which may alter motor function) should be measured postprandially.
Rapid transit is unlikely to result from an intestinal response to excessive secretion by the small bowel; in experimental cholera, there is prolongation rather than acceleration of small-bowel transit32. Purely secretory diarrhea is uncommon among patients with the carcinoid syndrome. Increased stool weight in 10 of 12 patients with the carcinoid syndrome and diarrhea in two reports were associated with increased stool fat, but the patients' stool weight rarely exceeded 700 g per day (range, 250 to 1407; mean, 514)2,33. These results suggest that, unless the transit time in the small bowel or colon was shortened, much of the volume load resulting from secretion by the small intestine would be reabsorbed, in view of the normal solute-coupled transport in the small bowel of patients with the carcinoid syndrome14 and the reabsorptive capacity of the proximal colon16. Indeed, when 500 ml of fluid is delivered to the ascending colon of normal subjects over a period of 36 minutes, there is accelerated emptying of liquids from this region, but no change in the overall colonic-transit time and no diarrhea34,35.
In summary, our findings support the hypothesis that the presumably hormonally mediated diarrhea in patients with the carcinoid syndrome and diarrhea involves gastrointestinal motor dysfunction28. Other hormones, such as gastrin, somatostatin, and vasoactive intestinal polypeptide, also alter gastrointestinal motor function; excessive circulating concentrations of these hormones may affect not only gastric or intestinal secretion but also gut motor function and contribute to the diarrhea associated with Zollinger-Ellison syndrome and medullary carcinoma of the thyroid36. The methods we used can be employed to study the motor function of the gut as well as provide quantifiable measures to assess novel pharmacotherapies in patients with hormonally induced diarrhea.
Supported in part by a Clinical Research Center grant (RR00585) from the National Institutes of Health, by a grant from Glaxo, Inc., by an American Gastroenterological Association/SmithKline Beecham award (to Dr. Camilleri), and by a grant (Oh53/1-2) from Deutsche Forschungsgemeinschaft (to Dr. von der Ohe).
We are indebted to Mr. Russell Hanson, Mr. Richard Tucker, and Dr. Mario Vassallo for technical support; to Dr. Gertrude M. Tyce for helping us set up the plasma serotonin assay; to Ms. Cindy Stanislav for assistance in the preparation of the manuscript; and to Dr. Lee A. Forstrom, who holds the investigational new drug licenses for radiolabeled Amberlite pellets, for his support.
Source Information
From the Gastroenterology Research Unit (M.R.O., M.C., G.M.T.) and the Division of Medical Oncology (L.K.K.), Mayo Clinic and Mayo Foundation, Rochester, Minn. Published in part in abstract form: Camilleri M, et al. Gut 1991;32:A1215 and von der Ohe MR, et al. Gastroenterology 1992;102:A251 and 1993;104:A595.
Address reprint requests to Dr. Camilleri at the Gastroenterology Research Unit, Mayo Clinic, Rochester, MN 55905.
References
| |||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
