SELECTED SUMMARIES ROBERT M. GLICKMAN

Selected Summaries Editor Columbia University College of Physicians & Surgeons New York, New York 10032

GASTROENTEROLOGY 1985;88:207-13

STAFF OF CONTRIBUTORS

Thomas A. Brasitus, Chicago, Ill. Richard J. Deckelbaum, Jerusalem, Israel Serge Eriinger, Paris, France

Sumner C. Kraft, Chicago, Ill. Seymour M. Sabesin, Memphis, Tenn. Melvin Schapiro, Los Angeles, Calif. Konrad Schulze-Delrieu, Iowa City, Iowa Joseph Sweeting, New York, N.Y. Robert S. Fisher, Philadelphia, Pa.

W. G. M. Hardison, San Diego, Calif. Khursheed Jeejeebhoy, Toronto, Canada Rayford S. Jones, Charlottesville, Va. Martin F. Kagnoff, San Diego, Calif.

Robert C. Kurtz, New York, N.Y. Jeffrey Lichtenstein, New York, N.Y. Richard P. MacDermott, St. Louis, Mo. James McManus, Temple, Tex. Esteban Mezey, Baltimore, Md.

Phillip P. Toskes, Gainesville, Fla. Martin H. Ulshen, Chapel Hill, N.C. Ernest Urban, San Antonio, Tex. Milton M. Weiser, Buffalo, N.Y.

Ann Ouyang, Philadelphia, Pa. David F. Ransohoff, Cleveland, Ohio

MASSIVE OBESITY, SLEEP APNEA, AND GASTRIC BYPASS

Peiser J, Lavie P, Ovnat A, Caruzi I (Department of Sur­gery, Soroka University Medical Center, Beer-Sheva, Isra­el) Sleep apnea syndrome in the morbidly obese as an indication for weight reduction surgery. Ann Surg 1984; 199(1):112-5 (January).

This study reports on 15 patients with sleep apnea syndrome (SAS) who demonstrated improvement in their apnea index after gastric bypass surgery for obesity. The authors comment that one-third of morbidly obese men being scheduled for gastric bypass surgery have symptoms suggestive of SAS and 90% have the diagnosis confirmed by polyhypnographic recordings. Patients were defined as having SASby an apneic index and by patient movements during sleep. The apneic index was reassessed 2 to 4 months after discharge and again 4 to 8 months after the first postoperative evaluation in 6 patients. The apneic index was improved in all patients, with 12 having the total number of apneic episodes at night returning to within normal limits. This improvement persisted in the 6 patients evaluated for a second time postoperatively. The patients had a mean ± SD weight of 141.5 ± 30.8 kg preoperatively and 106.4 ± 22.73 kg postoperatively. The arterial blood gases preoperatively were as follows: Peoz, range 32.0-49.0 Torr; Paz, range 63-92.4 Torr. These studies were not repeated postoperatively. The authors conclude that massive weight reduction by gastric bypass surgery can cure SAS in an unselected group of morbidly obese patients and that the findings in this study support the hypothesis that obesity is a primary rather than sec­ondary precipitating factor in the etiology of SAS in morbidly obese patients.

Comment. This is an interesting report suggesting a new indica­tion for gastric bypass surgery for weight reduction. There are a number of problems with the conclusions of this study which are related to the design of the study. Sleep apnea syndrome is diagnosed if, during 7 hours of nocturnal sleep, at least 30 apneic

episodes are observed both in rapid eye movement and nonrapiL eye movement sleep, some of which must appear in a repetitive sequence in nonrapid eye movement sleep. An apnea is defined as a cessation of airflow at the nose and mouth lasting at least 10 seconds (Ann Rev Med 1976;27:465-84). The patients studied in this report do fit the criteria for SAS. The authors indicate that this is an unselected group of patients. They state that one-third of morbidly obese male patients being scheduled for surgery have symptoms suggestive of SAS. There is no mention of the total number of male patients in the study. A truly unselected group would require all patients to undergo sleep studies to determine which subjects fit the criteria and to evaluate all subjects falling into the SAS category after surgery. The second conclusion drawn by the authors is that this study supports the hypothesis that obesity is a primary factor in the SAS. In large studies only 5% to 34% of SAS patients are obese (Ann Rev Med 1976;27:465-84). It has been described before that weight loss will improve the apneic index (Acta Med Scand 1960;167:343). In addition, sophis­ticated studies are now available to separate different types of apnea, such as central apnea, upper airway apnea, and mixed apnea (Brain Res 1966;2:167-86). In the patients with upper airway apnea, a tracheostomy has been successful in treating the nocturnal apneic episodes. This is true even for the massively obese patient (Am J Med 1974;56:531-9). Other authors have demonstrated a decreased hypoxic ventilatory drive in the pa­tients with obesity hypo ventilation (Am J Med 1975;59:343-7). Hypoxic ventilatory drive does not necessarily improve with weight reduction (Circulation 1967;36:11-258). In subjects with normal weight, a decreased hypoxic ventilatory drive does not usually produce hypoventilation, suggesting that the decreased hypoxic ventilatory drive combined with a resistance to breathing as produced by severe obesity may result in hypoventilation (Am Rev Respir Dis 1982;126:640-5). It should seem simplistic to conclude that improvement in the number of apneic episodes as a result of weight loss demonstrates that obesity is a primary factor. With these reservations aside, I think it is well established that in a massively obese patient with SAS, weight reduction may improve the patient's symptoms and that when considering an obese patient for gastric bypass, the presence of SAS would be an indication for surgery. In addition, patients undergoing surgery would be a group, which, if studied systematically for site of apnea and for hypoxia ventilatory drive, might provide an answer to the chicken or egg controversy that surrounds this issue.

A. OUYANG, M.D.

208 SELECTED SUMMARIES

VEILED MUCOSAL MONONUCLEAR CELLS IN CHRONIC INFLAMMATORY BOWEL DISEASE

Wilders MM, Drexhage HA, Kokje M, et a1. (Departments of Pathology and Gastroenterology, Free University, Amster­dam, and Department of Gastroenterology, University of Leiden, The Netherlands) Veiled cells in chronic idiopath­ic inflammatory bowel disease. Clin Exp Immunol 1984; 55:377-87 (February).

A recent Selected Summary (GASTROENTEROLOGY 1984;86:582) discussed the observation of nonlymphocy­tic dendritic or veiled mononuclear cells with long, active­ly moving cytoplasmic extensions among intestinal cell suspensions obtained from the rat, guinea pig, and pig (Immunology 1983;48:453-60). Similar veiled cells previ­ously had been described in the skin where they were believed to be involved in presenting antigen to T lympho­cytes (Cell Tissue Res 1979;202:407-30). In studies of sectioned rat Peyer's patches, the veiled cells (a) were present between and underneath the lymphoepithelium, in the upper part of the reticular area; (b) sent branches between the T cells in the T-dependent interfollicular area; and (c) often were seen directly beneath the M cells, the specialized surface epithelial cells that are considered involved in the sampling and uptake of antigens. The hypothesis was presented that intestinal veiled cells may pick up antigen from M cells, penetrate the basement membrane, and migrate below-forming an antigen-pre­senting-cell system in mucosa-associated lymphoid tissue that is comparable to that of the skin (Immunology 1983;48:461-7).

In this new study, veiled intestinal mucosal cells were identified and characterized in both cell suspensions and microscopic sections of surgical specimens and endoscop­ic biopsy specimens from 18 patients with chronic inflam­matory bowel disease (CIBD) (Crohn's disease, 9; ulcer­ative colitis, 7; indeterminate colitis, 2) and 18 control subjects ("normal" bowel in association with benign and malignant neoplasia, diverticulosis, or acute bacterial in­flammation). The cell suspensions were prepared by ex­tensive washing, collagenase dissociation, filtration, and related separation techniques. Characterization of the iso­lated cells involved phase-contrast microscopy, time-lapse cinematography, and electron microscopy after immune peroxidase staining; the tissue sections were studied with immune and enzyme histochemistry and by transmission electron microscopy.

Only two of the 18 control specimens appeared to contain a few veiled cells, but these cells easily could be found in 14 of the 18 CIBD specimens; no differences regarding the presence of veiled cells were observed be­tween Crohn's disease and ulcerative colitis. There were no clear-cut associations with sex, age, duration, localiza­tion, or activity of the bowel disease. The veiled cells in the CIBD specimens resembled those previously described in the rodent gut vis-a.-vis the pattern of motion, strong la­positivity, no or only weak acid phosphatase activity, and ultrastructure; the typical ultrastructural features included a lobulated horseshoe-shaped nucleus, electron-translu­cent cytoplasm with several cytoplasmic processes, and

GASTROENTEROLOGY Vol. 88, No.1, Part 1

many free ribosomes and mitochondria. However, there also were numerous la-positive veiled cells that contained increased numbers of lysosomes, residual bodies, and signs of cytophagocytosis-i.e., combining the characteris­tics of veiled cells with those of phagocytic macrophages. The la-positive veiled cells often appeared to be in juxta­position with one another as well as with lymphocytes; whereas prominent cellular aggregates were observed that contained intermingled, partially la-positive cell mem­branes. In addition, characteristic phagocytic macrophages were identified in both the CIBD and control gut tissues, representing virtually the only la-positive cells in the latter.

The data strongly suggest the presence of antigen-pre­senting veiled cells in the intestinal mucosa in CIBD but not in grossly normal control specimens. A possible expla­nation is that intense antigen handling takes place in the gut wall of CIBD patients in comparison with the normal bowel. However, classical macro phages expressing la anti­gens also have the capacity to present antigen, and inter­digitating cells can conceivably transform into phagocytic cells. Thus, the veiled cells in the gut in CIBD may exhibit both antigen-presenting and phagocytic functions. Func­tional tests of more highly purified cell types are needed to detect possible differences in antigen-presenting and relat­ed capacities. It also will be of interest to perform a prospective study on a larger group of CIBD patients as 3 of the 4 CIBD patients who were devoid of veiled cells clinically were in prolonged relapses that appeared resis­tant to prednisone therapy.

Comment. The fact that there appears to be a spectrum from typical veiled mononuclear cells to classical macro phages in the ileocolonic mucosa in patients with ulcerative colitis and Crohn's disease seems fitting inasmuch as these disorders are themselves part of a clinical spectrum. Wilders and colleagues further point out the heterogeneity of the reported histologic and immunologic features of patients with CIBD, especially as regards the disunity of the cellular infiltrates in the affected tissues. Although the present studies show a close relationship between veiled cells and lymphocytes in both ulcerative colitis and Crohn's disease, these authors have unpublished data that show a phenomenon resem­bling peripolesis (a process in which lymphocytes wander around larger target cells-in this case la-positive veiled cells) in 7 of 18 CIBD patients. This phenomenon was particularly easy to detect in cell suspensions from ulcerative colitis specimens. As peri­polesis involving veiled cells has been described earlier in asso­ciation with cell-mediated responses to contact antigens (Immu­nology 1983;49:415-22), it was suggested that this type of phenomenon might be playing a role in CIBD. Of course, the antigens involved in this process and hypothesized to be carried in the bowel wall by the veiled cells are at present unknown.

It is not proven that the main function of the gut veiled cells is comparable to that of the interdigitating cells present in the paracortical T-dependent areas of lymph nodes, Le., antigen presentation to the surrounding T cells (Cell Tissue Res 1979;202:407-30). In earlier studies of normal rabbit peripheral lymph nodes, for example, the veiled cells increased in number following antigenic stimulation (Anat Rec 1978;190:5-21). Fur­thermore, dendritic cells in the spleen are highly capable of physical association with responding T cells and may have major functions both in the afferent limb of the immune response and in the biology of the major histocompatibility complex (Immunol Rev 1980;53:127-47); while the injection of thoracic duct dendrit-

January 1985

ic cells has restored the ij'Ilmunogenicity of long-surviving renal allografts that"had been retransplanted into secondary recipients U Exp Med 1982;155:31-41).

Strong justification remains for documenting whether the veiled mucosal mononuclear cells , now described in both labora­tory animals and inflamed human tissues, are part of the intestinal antigen-presenting system. Increasing evidence in this direction would have important implications with respect to the chroni­cally elusive antigenic component(s) in ulc'erative colitis and Crohn's disease.

S. C. KRAFT, M.D.

YOGURT: A MAGIC POTION FOR THE LACTOSE INTOLERANT

Kolors IC, Levitt MD, Aouji M, et 01. (Research Service, Veterans Administration Medical Center and Departments of Medicine and Food Science and Nutrition, University of Minne;;ota, Minneapolis and st. Paul, Minn.) Yogurt-an autodigesting source of lactose. N Engl J Med 1984;310:1-3 Uanuary 5) . .

Why do a number of lactase-deficient population groups consume large amounts of yogurt (or other cultured milk products) but seldom ingest appreciable quantities of unmodified milk? The authors sought to answer this question by exploring whether lactose present in yogurt is absorbed better than lactose from milk. Ten healthy sub­jects aged 20-28 years were studied. These subjects were designated lactose intolerant on the basis of excreting more than 20 parts per million hydrogen in their breaths after ingesting 20 g of lactose. For the study, end-alveolar breath samples were obtained in the fasting state and then for 8 hours after ingestion of the following substrates: 10 g of lactulose (a disaccharide not hydrolyzed in human small bowel), 20 g of lactose, 400 ml of milk (containing 18 g of lactose), and 440 g and 270 g of a commercial unflavored yogurt (containing 18 g and 11 g of lactose, respectively). Three subjects were also intubated with a double-lumen pancreatic drainage tube so that the distal opening was in the duodenum. The subjects ingested 350 g of yogurt and several duodenal (and gastric) samples were obtained over the ensuing 90 minutes. The duodenal samples were analyzed for lactose concentration and lac­tase activity. In vitro samples of ypgurt were incubated at 4°C and pH 4.6 (the natural pH of yogurt). Similar samples were sonicated and then incubated at 37°C, pH 7.0. Hydro­lysis of lactose was measured by the appearance of galac­tose. The results clearly show that ingestion of 18 g of lactose in yogurt resulted in peak breath hydrogen concen­trations that Were only about one-third as great as with the ingestion of 20 g of lactose alone. Ingestion of the smaller amount of yogurt showed a still lower breath hydrogen peak, whereas after ingestion of 10 g of lactulose the increase in breath hydrogen was comparable to the effects of 20 g of lactose. Diarrhea and flatulence were reported by 80% of the subjects ingesting milk but by only 20% after ingesting yogurt. Assays of duodenal aspirates for lactase activity in the three intubated subjects showed negligible lactase activity before but appreciable activity for at least 1 hour after ingestion of yogurt. The aspirates over the first

SELECTED SUMMARIES 209

hour contained sufficient lactase activity to digest 50%-100% of the measured lactose in the aspirate in 4 hours. In vitro incubation of yogurt at 4°C, pH 4.6, resulted in a negligible appearance of galactose, whereas in sonicated yogurt incubated at 37°C, pH 7.0, there was a significant amount of galactose. The authors conclude that bacterial lactase survives passage through the stomach and that it substitutes for the lack of endogenous lactase in lactase­deficient persons. They speculate that this autodigesting feature accounts for the popularity of yogurt in Middle Eastern countries where lactase deficiency is prevalent.

Comment. The above report created considerable interest. Its publication was accompanied by an editorial (N Engl J Med 1984;310:42-3) and a further commentary has also appeared (Nutr Rev 1984;42:2i6-8). A closely related study was published al­most simultaneously (J Dairy Sci 1984;67:1-6). That ~tudy, also utilizing breath hydrogen excretion as a measure of lactose malabsorption, compared ingestion of heated and unheated cul­tured yogurt in subjects who had difficulty digesting lactose. I will comment further on this.

Historically, cultured dairy products have long been touted as having therapeutic value in alleviating human and animal disor­ders. An early postulate by Metchiilkoff in 1~08 theorized that toxin production by putrefactive bacteria in the human intestine was suppressed. The longevity of Bulgarians was said to be in part due to their ingesting large quantities of Bulgaricus milk (J Dairy Sci 1979;62:1685-94). In 1970, Baer (Soc BioI 1970;17:143) sug­gested, but offered no evidence, that lactose-intolerant individuals might be able to consume yogurt without developing symptoms normally associated with lactose malabsorption. Four years later Gallagher et a!. described 3 lactase-deficient patients who tolerat­ed fermented dairy products without symptoms whereas nonfer­men ted dairy products caused moderate to severe symptoms (J Am Dietet Assoc 1974;65:418-9). There have been several reports about yogurt and lactase since then (J Dairy Sci 1976;59:601-6 and 2031-5, 1982;65:346-52).

Primary lactase deficien.cy is the norm for the majority of the world's adult population and indeed is the norm for the vast majority of the adult mammalian kingdom where capability of digesting lactose declines sharply in the period beyond infancy. All present populations with intestinal mucosal lactase persisting into adulthood seem to have originated in southwest Asia where 5000 to 8000 years ago mankind began milking domestic sheep, goats, cattle, and camels (Lancet 1975;ii:910-1, November 8) . At that time, the mutation may have had biologic fitness and survival value. The passage of about 400 generations since then is said to account for the presently observed phenotype frequency (Gastro­enterology 1972;63:524-6) .

Most Americans believe that milk is an essential part of a normal diet. This national love pf milk (or is it an obsession?) is of course aided and abetted by the dairy industry. In 1979, Ameri­cans are reported to have consumed 256 pounds of milk and milk products per person! (Current Contents 1980;#49:5-8, December 8). For the adult who is lactose intolerant, help is at hand: low intestinal mucosal lactase levels can be overcome! Bacterial fermentation of milk will utilize some of the lactose in milk with production of lactic acid. As the pH decreases, fermentation will slow and eventually stop. The low pH, as well as possible production of antimicrobial agents during fermentation , inhibits a wide spectrum of food spoilage organisms , a feature particularly useful in the absence of refrigeration. Examples of fermented milk products besides yogurt are acidophilus milk, bifidus milk, but­termilk, kefir, ropy milk, and quark (J Dairy Sci 1982;65:346-52, Arch Latinoam Nutr 1983;33:247-56). Instead of bacterial fermen­tation, acids may be added to milk to precipitate the protein curd

210 SELECTED SUMMARIES

which can then easily be separated from the whey before con­sumption. At least cottage cheese, sour cream, and yogurt may be made by adding acidulants as well as by fermentation (J Dairy Sci 1976;59:2031-5). Lime juice is used in Mexico and baobab juice in Africa for this purpose (Lancet 1975;ii:910-1, November 8).

There are no precise reGords of where or when yogurt was first made. Legend has it that it was brought down in a pot by an angel. Other stories suggest that the people in the Middle East acciden­tally discovered yogurt when milk was left too long in a warm place. Whatever its origin, yogurt made by fermentation is the product of milk inoculated with both Lactobacillus bulgaricus and Streptococcus thermophil us. Fermentation reduces lactose by 20%-30%, the lactic acid providing the flavor and thickening the milk. In present day manufacture of commercial cultured yogurt, the texture is improved by increasing milk solids by about 30% either by dehydration or by the addition of dried milk solids before fermentation. Therefore, the lactose concentration in com­mercial yogurt is about the same as in whole milk. Various flavorings, colorings, sweeteners, and stabilizers are also added. Other differences in the various yogurts refer to the amount of fat-whole milk yogurt contains 3.25% milk fat, low fat yogurt has 0.5%-2%, and nonfat yogurt has <0.5% fat. To improve keeping qualities, i.e., extend shelf life, some commercial yogurts are pasteurized after fermentation. This kills the culture microflo­ra, destroys their endogenous lactase, and thereby effectively nullifies benefits for the lactose intolerant (J Dairy Sci 1984;67:1-6).

Goodenough and Kleyn have shown there is survival of the culture organisms in the in vivo rat small intestine for up to 3 hours after feeding natural yogurt (J Dairy Sci 1976;59:601-6). Both culture organisms contain endogenous lactase and at least the lactase of S. thermophilus is inducible (J Dairy Sci 1976;59:2031-5). In their studies, Kolars and coworkers showed that endogenous bacterial lactase in yogurt may be released by disrupting the bacterial wall with sonication. Bile also enhances the lactase activity of yogurt (J Dairy Sci 1984;67:1-6) and this is the likely mechanism for lactase release from yogurt in the in vivo small intestinal lumen.

Several additional points are relevant. First, a person who is lactase deficient need not necessarily be lactose intolerant, i.e., the two terms are not synonymous. Intolerance refers to the gas, bloating, cramping, borborygmi, and diarrhea that may ensue after ingestion of lactose. There is great variability in the extent and severity of symptoms amongst individuals challenged with a lactose load (Lactose digestion: clinical and llutritional implica­tions. Paige OM, Bayless TM, eds. Baltimore: Johns Hopkins University Press, 1981:124-33). The nutritional value of lactose­containing foods is largely retained by lactase-deficient and lactose-intolerant persons. Much of the unabsorbed lactose is salvaged in the colon. Volatile fatty acids, which are a principal product of colonic bacterial fermentation, are readily absorbed (J Clin Invest 1976;57:1158-64, Gastroenterology 1980;78:444-7). Most lactose-intolerant individuals learn to titrate their intake of lactose-containing foods to avoid symptoms. Thus, the nutritional consequences of lactose intolerance in adults are probably mini­mal except for calcium. Dairy products are our main dietary source of calcillm. There is epidemiologic evidence of a relation­ship between calcium intake, osteoporosis, and bone fracture rate (Lancet 1983;ii:1181-5, November 19). Another epidemiologic relationship, where there is so far no clear consensus, concerns dietary calcium and magnesium intakes in relation to sodium intake and hypertension (Nutr Rev 1984;42:205-13 and 235-6). Thus, yogurt and other fermented dairy products that retain considerable bacterial lactase (i.e., are not pasteurized after fer­mentation) may have a useful role for those who severely limit their intake of dairy products because of lactose intolerance.

Footnote: Apart from fermented dairy products with endoge­nous lactase, lactase offungal origins is now also available both in a liquid (LactAid, LactAid Inc., Pleasantville, N.J.) and in capsules

GASTROENTEROLOGY Vol. 88, No.1, Part 1

(Lactrase, Kremers-Urban Company, Milwaukee, Wis.). The liquid is used to treat milk which is then kept in a refrigerator for 24 hours for hydrolysis of much of its lactose to the component monosaccharides glucose and galactose. This low-lactose milk will taste sweeter than untreated milk because of the relatively sweeter taste of glucose. Alternatively, one or more capsules of lactase can be taken orally just before and during ingestion of dairy products or foods containing added milk solids. As with yogurt, hydrolysis of lactose occurs intraluminally during passage of the food. For obvious reasons, a better effect is likely if the food or drink is not ingested ice-cold. It seems safe to titrate the oral dose of lactase upwards until symptoms are controlled.

E. URBAN, M.B.B.S.

HIGHWAYS IN THE CELL

Geuze HI, Slot IW, Strous GIAM, et a1. (Laboratory for Cell Biology, University of Utrecht, The Netherlands; Chester Beatty Research Institute, Sutton, Surrey, England; Physio­logisch-Chemisches Institut, University of Munster, Mun­ster, F.R.G.; and Division of Pediatric Hematology-Oncol­ogy, Children's Hospital, Boston, Mass.) Intracellular receptor sorting during endocytosis: comparative immun­oelectron microscopy of multiple receptors in rat liver. Cell 1984;37:195-204.

The authors have used a clever method of "visual" dual­labeling to observe simultaneously the pathways of two different receptor-ligand complexes through the rat liver cell. They chose three different receptor systems for study that reasonably would be expected to take different path­ways after endocytosis. The asialoglycoprotein receptor system (ASGP-R) can be expected to uncouple early with delivery of li~and to lysosomes and receptor to the plasma membrane. The mannose-6-phosphate receptor (MP' R) also delivers its ligand to the lysosomes but the site of uncoupling is unknown. The immoglobulin A system (IgA-R) remains intact as it traverses the cell and exits via the canaliculi. Antibodies to each of the three systems were prepared by binding colloidal gold particles (visible by electron microscopy) to purified antireceptor antibod­ies. When the gold-antibody complexes were then layered over tissue sections, the antibody would bind to its recep­tor and the gold, visible by electron microscopy, would likewise be bound. By using two different sizes of gold particles, the authors could visualize the locations of two different receptors in a single tissue section. They con­firmed earlier work on the pathways of the ASGP-R and IgA-R systems. The distribution of the MP-R system was similar to that of the ASGP-R system. The most significant new data came from the IgA-R and ASGP-R co localization experiment. The two systems were mixed together on the plasma membrane and coated pits. However, by the time these two systems had reached the compartment for un­coupling of receptor aqd ligand (CURL) they had come to occupy separate microdomains. The IgA-R was largely localized to vesicle with a dense material at the luminal side of their limiting membrane. IgA labeling was fre­quently seen in continuity with bile canaliculi but was not identified, as ASGP was, in the lysosomes.

Comment. Reading this paper will not help you to better treat ulcerative colitis, nor will it allow you to heal that refractory

January 1985

duodenal ulcer in your problem patient. But insofar as knowledge of disease is based on knowledge of normal cell and organ function, time spent reading the article is a good investment. First, it is exciting to gain insjghts into the methods that have been devised to probe the workings in the gut of an object so small that its insides can be appreciated only by electron microscopy. Moreover the probe used is direct and comprehensible. It is not a diffractive or resonating pattern which requires a physicist to understand and interpret. The probe is a small black sphere clearly visible by electron microscopy. It is so clearly visible in fact that two different sizes of probe are easy to distinguish (4 vs. 8 mm). It is a vast improvement over older electron-dense labels such as ferritin. The specificity of targeting is also remarkable. Although not quite so new in concept, antibody labeling has been made more specific by affinity chromatography techniques for antibody purification and by the use of staphylococcal protein A which binds the colloidal gold avidly to antibody. The data are simple to understand but the technology that has gone into production of the data is complex and the fruits of many years' work by many groups in basic cell biology and biochemistry.

Second, it is exciting to discover what these methods tell us: how the cell works. Accumulated data suggest that substances taken up by the cell via specific membrane receptors traverse many crossroads as they travel the cell. This work demonstrates that different substances may be fellow-travelers early in a journey only to become separated at a critical junction. Although the present work concerns itself with exogenous ligands, it is true that substances originating within the cell must also take specific pathways either to plasma membrane or other subcellular organ­elles. In fact, the MP-R is probably more concerned with intracel­lular traffic than with receptor-mediated endocytosis (J Chem Biochem 1982;19;67). As exciting and new as these data seem, however, they are only the beginning. Once pathways are mapped, the important questions will deal with what signals regulate the traffic. Answers to such questions may unlock the keys to exquisitely directed subcellular manipulation and thera­py. Having demonstrated subcellular highways we must now seek to regulate the traffic signals.

W. C. M. HARDISON, M.D.

A FEAST OF FIBER

Jacobs LH, Lupton JH (Department of Internal Medicine, University of California, Davis, Calif.). Effect of dietary fibers on rat large bowel mucosal growth and cell prolifera­tion. Am J Physiol 1984;246 (Gastrointest Liver Physiol . 9):G378-85 (April).

In this study, the mucosa of the large bowel was exam­ined in rats after 4 weeks of feeding groups of 10 animals either a fiber-free diet (control group) or the fiber-free diet to which either 20% oat bran, 10% pectin, or 10% guar had been added. Rats were given intraperitoneal [3Hlthymi­dine and killed either 1 hour or 23 hours later. Animals were killed at the same time of day to avoid artifacts due to diurnal variations. The cecum and prQximal and distal colon were assayed for intestinal mucosal mass, DNA and RNA content, and [3Hlthymidine incorporation into DNA. In addition, autoradiography and histologic analyses of crypt column height and circumference were carried out on segments from each of the three regions of large bowel. The size of the proliferative and non proliferative compart­ments of crypts in each anatomic region was calculated and cell migration rates were estimated by comparing data

SELECTED SUMMARIES 211

from animals killed 1 hour and 23 hours after injection of [3Hlthymidine. Assuming a constant velocity, the estimat­ed migration rate was used to determine how many hours it would take for the leading edge of labeled cells to reach the crypt mouth. Overall, the data show that guar pro­duced the greatest stimulation of mucosal cell growth and proliferation, bran was the least effective, and pectin was intermediate. The largest effect was usually found in cecum and changes decreased distally. Specifically, in guar-fed animals, cecum showed the gre\ltest increase of mucosal weight, DNA, and RNA, whereas bran had no effect on these parameters in any large bowel segment. MUcosal DNA synthesis was variably affected by the three fibers. In cecum, guar increased cell numbers rather than mitotic activity per cell whereas in the pectin-fed animals [3Hlthymidine incorporation decreased per microgram DNA but not per microgram mucosal weight. This suggests that the ratio of proliferative to ponproliferative compart­ment decreased compared with the fiber-free control group. In proximal colon, pectin did not alter [3Hlthymi­dine incorporation into DNA but bran and guar did . In distal colon, [3Hlthymidine incorporation into DNA was unchanged by any of the fiber diets. Crypt morphometrics showed that in cecum guar feeding resulted in crypt cell hyperplasia (deeper crypts). pectin decreased crypt cir­cumference, and bran again had no effect. In proximal colon, all three fibers decreased the total number of cells per crypt but, with constant or increased DNA levels, there must have occurred an increase in the number of crypts per unit lengths of intestine. In distal colon there were no changes. None of the three fibers altered the labeling indices in cecum Or proximal colon, but in distal colon guar resulted in the greatest changes. Analysis of migration rates showed that it took longer for the leading edge of labeled cells to reach the mouth of crypts in all fiber-fed animals compared with control animals fed the fiber-free diet.

CQmment. A decreased fiber intake concomitant with ingestion of increasingly refined foods has resulted in modern man being lampooned as overweight, toothless, and constipated. The intake of crude fiber in the American diet has decreased almost 30% since the beginning of this century (Am J Clin Nutr 1978;31:1510-4) . Perhaps because of this , dietary fiber has become the "in thing. " What is dietary fiber? Most simply it is the supporting walls of plant cells. Because it provides protection, it is also tough and mammalian small intestine cannot digest it. This last fact was used by Trowell for a broadly based definition of dietary fiber-it is the residue of plant food that is resistant to hydrolysis by mammalian alimentary enzymes (AmJ Clin Nutr 1976;29:417-27; Proc NutrSoc 1984;43:25-33). Fiber can be divided into two chemical classes: (a) non-a-glucan polysaccharides (cellulose, hemicellulose, pectins) and (b) lignins. Physiologic and physical properties of dietary fiber include effects on texture and palatabil­ity of foods, satiety effects, effects on intestinal transit time, intraluminal pressure, and release of gastrointestinal hormones as well as water-holding capacity and cation-exchange and adsorp­ti ve capabilities (Physiology of the gastrointestinal tract, Johnson LR, ed. New York: Raven, 1981 :1291-9; J Lipid Res 1982;23:221-42). Fiber is also fermentable by bacteria. This occurs not only in the large bowel of carnivores and humans but also in the rumen and large bowel of herbivores. Similarities in secretory and absorptive capabilities and in microbial digestion of fiber in rumen and large bowel have recently been pointed out. In both organs, end products of anaerobic carbohydrate fermentation are

212 SELECTED SUMMARIES

volatile short-chain fatty acids (acetic, propionic, butyric) which are absorbed at equivalent rates. Digesta pH are maintained by the release of bicarbonate from mucosa into the lumen together with the buffering action of either saliva or ileal effluent (Proc Nutr Soc 1984;43:13-2~). So much for the simplistic view that dietary fiber only provides unabsorbed bulk for the diet that aids elimination of waste products of digestion.

Dr. Jacobs has previously described the effects of oat bran, pectin, and guar compared with total fiber deprivation on muco­sal growth and cell proliferation in the small intestine of the rat (Am J CUn Nutr 1983;37:954-60). In that study, like the present one directed at large bowel morphometrics, guar produced the greatest effects, pectin was intermediate, and the results of adding oat bran differed little from animals fed a fiber-free diet. Thus, the more soluble forms of fiber seem to result in greater mucosal modulations in both small and large bowel. In the maintenance of small bowel mucosal mass, three major and two minor factors have been identified: luminal nutrition, pancreatic-biliary secre­tions, enteric hormones; mucosal blood supply and neural effects. No single factor can fully explain all small bowel mucosal adaptive responses and presumably several factors contribute. Although several critical reviews of small bowel mucosal adapta­tion and its regulation have been published (N Engl J Med 1978;298:1393-402 and 1444-50; ;'\dv Intern Med 1980;26:265-91; Scand J GastroenteroI1982;17(Suppl 74):53-74), much less is known about adaptive responses of large bowel mucosa. Consid­erable material about adaptive responses of large bowel mucosa was presented in a recent Selected Summary (Gastroenterology 1984;86:208-9) which commented on effects of wheat bran on mucosal proliferation in small and large intestine. Briefly, disten­tion per se plays little if any role in large bowel mucosal hyperplasia. The role, if any, of pancreatico-biliary secretions in the maintenance of large bowel mass is unknown. Dietary fiber inf!.uences the serum concentration of several gastrointestinal hormones (gastrin, glucagon, GIP) (Colon and nutrition, Kasper H, Goebell H, eds. Hingham, Mass.: MTP Press, 1982:71-6). The presence of luminal nutrient is likely to be important for mainte­nance of large bowel mucosa. Whether this is unabsorbed nutrient from ileal effluent or undigested fiber is immaterial. As already mentioned earlier, anaerobic bacterial fermentation of fiber in large bowel results in the production of volatile fatty acids. Most of the fermentation occurs in cecum. Lipids have been shown to be powerful stimulants of mucosal growth in remaining small bowel of rats after intestinal resection. Small quantities of long­chain triglycerides promoted small bowel mucosal growth more than protein or polysaccharide. The quantity of oil also played a role, as less oil gave less adaptation. Medium-chain triglycerides were less effective than long-chain triglycerides in stimulating small bowel mucosal growth [Gastroenterology 1978;74:1070 (abstr); Mechanisms of intestinal adaptation, Robinson JWL, Dowling RH, Riecken EO, eds. Hingham, Mass.: MTP Press, 1982:175-85]. Free fatty acids should have similar or perhaps even greater mucosal growth effects but to my knowledge no data have so far been published. Also, I am not aware of any data concerning effects of short-chain triglycerides or short-chain fatty acids on adaptive increases of either small bowel or large bowel mucosal mass. In addition, we don't know why different fibers elicit different proliferative responses such as illustrated in the paper by Jacobs and Lupton. Obviously much mote work remains to be done. Dr. Jacobs stated recently that individual dietary fibers can no longer be lumped together as "bulk" or "fiber" (Gastroen­terology Hl84;86:372-3). I agree wholeheartedly. Detailed studies should ultimately lead to a rational use of dietary fiber in both health and disease (Gastroenterology 1983;85:220-1).

E. URBAN, M.B.B.S.

GASTROENTEROLOGY Vol. 88, No.1, Part 1

MARKERS FOR COLONIC NEOPLASIA

Luk CD, Baylin SB (Oncology Center and Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Md.) Ornithine decarboxylase as a biologic marker in familial colonic polyposis. N Engl J Med 1984; 311:80-3 (July).

Patients with familial polyposis eventually develop co­lonic cancer. In those family members who carry the genotype without evidence of polyposis, there is a need for careful screening. Radiologic and endoscopic techniques only detect the condition after the macroscopic develop­ment of polyps. Cell kinetic measurements on mucosal biopsy specimens have demonstrated evidence of a hyper­proliferative state. Ornithine decarboxylase (ODC) is the rate-limiting enzyme in polyamine synthesis and appears to be a necessary step in rapid tissue growth. For these reasons ODC activity was measured in specimens of colon­ic mucosa and polyps from patients with polyposis and at­risk family members.

Three groups of subjects were studied: (a) normal con­trols, (b) unaffected first-degree family members with a 50% probability of carrying the genotype, and (c) affected family members with colonic polyposis. Biopsy specimens were separated into flat normal-appearing mucosa, adeno­matous polyps, and dysplastic polyps. No polyps con­tained invasive carcinoma.

Mean ODC activity was lowest in the specimens from normal controls and highest in the dysplastic polyps with intermediate but increasing levels in the flat mucosa and adenomatous polyps, respectively, from the polyposis subjects. This suggested a correlation between increased proliferative activity and the potential for tumor develop­ment. Using an llPper limit of normal for ODC activity of 2.5 nmol/mg . h, specificity was 100% and sensitivity 85%. In the group of unaffected but at-risk patients. 45% of ODC values were elevated above control levels. The authors concluded that measurement of colonic mucosal ODC activity may identify clinically normal family members who carry the genotype for familial polyposis.

Comment. Labeling of colonic epithelial cells by tritiated thymi­dine labeling has been previously used to successfully identify individuals with genetic susceptibility to colonic cancer (Cancer Res 1983;43:1899-904). Radioautographic characterization of co­lonic cell proliferative activity could therefore probably be regard­ed as the "gold standard" to which all newer measurements should be compared. Unfortunately, because radioautographic studies were not carried out in this study, it is not possible to determine how measurements of ODC activity compare with thymidine labeling or whether these two measurements comple­ment each other. Although both types of measurement must be performed on mucosal biopsy specimens, the ready availability of colonoscopy and the safety of biopsy make tissue sampling a feasible approach to widespread screening of high-risk groups.

Measurement of polyamines in pathological tissues and fluids from cancer patients is not, however, specific inasmuch as a rise in tissue polyamine levels is common to rapidly growing tissues. A recent report (Cancer Res 1984;44:845-7) has described a selective elevation of N"-acetylspermidine in colorectal cancers that was not present in benign adenomas. This same paper states that preliminary studies in patients with juvenile polyps, lym-

January 1985

phoid polyps, or ulcerative colitis (frequently a hyperproliferative condition) showed no elevation of N'-acetylspermidine. It will be of great interest to examine the sensitivity and specificity of ODC measurements in other premalignant conditions such as ulcer­ative colitis. In addition, long-term prospective studies are now needed to determine how reliable elevated ODC levels are in predicting the future development of polyps in the clinically normal kindred of familial polyposis patients.

A further possible application of the observed elevation in polyamines and ODC activity in colonic neoplasia is the develop­ment of therapeutic agents, such as 2-difluoromethylornithine, an irreversible inhibitor of ODC activity. This agent has been used to suppress human tumor cell growth both in vitro and in vivo (Am J PhysioI1982;243:C212-21) and to inhibit experimental induction of colon cancer in mice (Cancer Res 1983;43:2545-9). If elevation of ODC activity is an obligatory step in polyp progression, then the use of 2-difluoromethylornithine or similar agents may have important application in chemotherapy and chemoprevention of polyps and cancer in affected individuals.

L. R. JACOBS, M.D.

REPLY TO SELECTED SUMMARY: MALNUTRITION IN ALCOHOLIC HEP A TITIS (GASTROENTEROLOGY 1984;87:1396)

Thank you for providing us an opportunity to reply to the summary by Dr. Mezey. We should like to add several additional comments. After reviewing the two articles cited (Am J Clin Nutr 1983;38:469-73 and 849-59), it is not possible to explain with complete certainty their observations that alcoholic hepatitis and cirrhosis were associated with an increased percent of ideal body weight and normal or increased fat stores. These observations are quite different from ours. One can only examine differ­ences in patient population and methods of assessment, for a possible explanation.

As pointed out by the reviewer, sex and country of origin were quite different. Mean alcohol intake in our patients was comparable to the other two studies but

SELECTED SUMMARIES 213

represented a slightly smaller percent of total caloric intake (45% vs. 52% and 65%). In the two studies cited, not all the patients had alcoholic hepatitis. Only 10 of 84 Chilean patients and 18 of 30 patients from Glasgow had alcoholic hepatitis. In addition, our patients were appar­ently much sicker clinically. The mean serum bilirubin in our patients was 164.3 /-LmollL vs. 58.4 /-LmollL in the patients from Glasgow (no data are given on the Chilean patients). In our patients ascites was observed in 61.6%, and hepatic encephalopathy in 44.9% with an overall mortality of 23.8%. Mills reported a 13.3% mortality, and Bunout did not give mortality data at all.

The clinical severity of alcoholic hepatitis is an impor­tant factor since in our study clinically mild disease had a significantly higher percent ideal body weight than severe disease (98.4% vs. 89.3%, p < 0.01). Furthermore, in our study normal or slightly higher ideal body weights were observed in 48% of mild cases of alcoholic hepatitis. Inasmuch as the other studies deal with a small number of patients with alcoholic hepatitis (10 and 18 patients) who have primarily mild disease, we believe this does not represent the complete nutritional picture for all severities of alcoholic hepatitis. Our study evaluated 156 mild cases, 108 moderate cases, and 99 severe cases.

Another point of importance is that percent of ideal body weight must be based on dry weight to reflect nutrition. Failure to correct for fluid weight in the pres­ence of ascites would result in enormously high body weights. We attempted to correct for the presence of ascites and observed that, on average, tense ascites repre­sented 12.5 kg of water, moderate ascites 6.1 kg, and minimal ascites 2.1 kg. No mention is made in either of these studies regarding this obvious source of error.

We believe that the difference in severity of the alcoholic hepatitis in a relatively small number of patients of differ­ent ethnic origins, as well as the possible failure to correct for fluid weight, may be responsible for most of our observed differences.

CHARLES L. MENDENHALL, M.D., Ph.D.

ROBERT E. WEESNER, M.D.