- Email: [email protected]
Maltose absorption as an indicator of small-intestinal allograft rejection
JOURNAL
OF SURGICAL
RESEARCH
Maltose Absorption
37,75-82
(I 984)
as an indicator of Small-Intestinal
Allograft
TIMOTHY R. BILLIAR, M.D., CARLOS GARBERDGLIO, ANDWOLFGANG H. SCHRAUT,M.D.*
Rejection’
M.D.,
Department of Surgery, University of Chicago Hospitals and Clinics, 5841 South Maryland Avenue, Chicago, Illinois 60637
Presented at the Annual Meeting of the Association for Academic Surgery, Syracuse, New York, November 2-5, 1983 Maltose and lactose absorption, which require an intact brush border for breakdown and absorption as glucose, was evaluated as a function test to monitor the integrity of the small-bowel graft. Using the rat model of accessory small-bowel transplantation, absorption tests (in the form of an oral glucose tolerance test) were performed on iso- and allografts with either portal (PP-A) or caval venous drainage (PC-A). In isografts the absorption of maltose was found to be reproducible and not influenced by the type of venous drainage. This was not the case with the use of lactose which thus was not studied further. Allografts with PC-A demonstrated a reduction in their capacity for maltose absorption on the fifth postoperative day, as the glucose peak at 30 min (T30) was significantly blunted in comparison to that for isografts with PC-A (167 mg% 2 12 vs 204 mg% 5 8). Functional impairment preceded histologic changeswhich did not arise before the sixth-to-seventh postoperative day in rats with PC-A. Allografis with PP-A absorbed maltose on the fifth postoperative day nearly as effectively as did isografts (T30 min: 185 mg% ? 14 vs 213 mg% + 8). By the ninth postoperative day, the serum glucose curve after maltose administration was flattened for grafts with PC-A (T30 min: I37 mg% f 1I) which were rejected acutely (host’s death) after 11.8 days + 0.45. A similar impairment of maltose absorption was not seen in the PP-A group (chronic graft rejection after 22.8 days + 1.8) until the 15th postoperative day. These results suggestthat maltose absorption becomes impaired before the clinical appearance of rejection and its histologic alterations. Tbe maltose absorption test effectively predicts rejection and monitors the function of the intestinal graft in this animal model.
elusively. Furthermore, this method of monitoring requires a cutaneous enterostoma, The expectant monitoring of small intessince orthotopically placed grails are relatively tinal allografts for rejection presentsa problem inaccessible for frequent biopsy. In addition, not seen with other types of organ transplana small-bowel graft that appears morphologtation becausea simple, reproducible function ically to be intact does not necessarily have test is unavailable-a test that would reflect functional integrity. functional and thus morphologic integrity of We postulate that intestinal absorption of the allog&. At present, histologic evaluation the disaccharides, maltose and lactose, may of serial mucosal biopsy specimensis the most provide a sensitive function test. This test recommon method of following the rejection quires the enzymes (disaccharidases) of the process. However, at least in canine and rat mucosal brush border to be intact for the intestinal allografts, the histologic alterations breakdown of maltose and lactose into glucose that occur during rejection are known to be and fructose and for the active absorption of variable in both distribution and intensity these monosaccharides. In this study on a rat [l-3]. This indicates that random biopsies may model of accessory small-bowel transplantafail to reflect the status of the allograft contion, we measured the absorption of maltose ’ Supported by the National Institutes of Arthritis, Me- and lactose in the form of a glucose tolerance test. We sought to determine whether this test tabolism and Digestive Diseases(Grant AM27332). ’ To whom requests for reprints should be addressed. can be used as a method for predicting and INTRODUCTION
75
0022-4804/84 $1.50 Copyright Q 1984 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984
16
monitoring the rejection of small-bowel allografts as it becomes evident from histologic and serum-immunologic studies. MATERIALS
AND METHODS
We used rats of the inbred Lewis (LEW) and Lewis X Brown-Norway F- 1 hybrid (LBN) strains, weighing between 180 and 250 g. Absorption studies were carried out in the normal (nontransplanted, in situ) small bowel, in insografts, and in allogrtis (Table 1). Operative procedures. For absorption studies on the nontransplanted small bowel, LEW rats were used. Under chloral hydrate [3 mg/gm body wt (SW)] anesthesia, a Silastic catheter was placed via an enterostomy into the proximal jejunum of the fasted animal. The bowel was irrigated with 0.45% saline solution and then was occluded proximally (duodenum) and distally (ileocecal junction). This was followed by instillation of the test solution. After each absorption test, the enterostomy site was closed and the animal allowed to recuperate. Three tests per animal were undertaken at I- to 2-week intervals. The entire small bowel was transplanted as an accessory graft according to a procedure described previously [4]. Isografts and allografts, the latter with the LEB - LEW strain TABLE I Experimental group
No. of rats
Normal, nontransplanted small bowel
4
Small-bowel isografts with Portocaval anastomosis (PC-A) Portoportal anastomosis (PP-A)
4 5
Total Small-bowel allografts with Portocaval anastomosis (PC-A) Portoportal anastomosis (PP-A) Total
(9) 24” 12” (36)’
a Absorption studies were carried out on only six rats in each group. ‘Excludes nine rats that died during the absorption tests.
combination, were transplanted with provision made for either portal (portoportal anastomosis, PP-A) or systemic (portocaval) anastomosis, PC-A) venous drainage. Absorption tests.Baseline absorption studies were performed with normal saline, glucose, lactose, and maltose in the small bowel of normal rats (N = 4). Glucose (1 mg/g BW), lactose (1 mg/g BW), or maltose (0.5 mg/g BW) was dissolved in 2.5 ml of 0.45% saline solution and instilled into the bowel. Serum glucose levels were measured with a Beckman glucoseanalyzer at 5- to 15-min intervals until the glucose level returned to the previously recorded baseline. Since the initial measurements showed that the absorption of lactose was highly variable from test to test, we limited the absorption studies in rats with iso- and allografts to the use of maltose as substrate. In rats with isografts the maltose absorption (MAb) test was performed repeatedly (three to four times) on random days from the seventh postoperative day up to 3 months postoperatively. In 12 of the 40 rats with allografts, the test was undertaken successively on the 5th, 9th, 12th, and 15th postoperative days, unless graft rejection prevented the next test. On the day before and the day of the MAb test, the graft was irrigated clear with 5- 10 cc of normal saline. The maltose solution was infused into the proximal stoma, and both the proximal and distal stoma were then occluded with balloon catheters. Histologic examination. Upon death or sacrifice of the rats, all grafts were fixed in formalin, and sections were prepared (H&E stain) for light microscopy. Hemagglutinin antibody assay. In order to have a measure of the host’s immune reaction to the small-bowel allograft, we measuredantidonor hemagglutinin titers. Washed LBN erythrocytes were incubated for 45 min with normal (LEW) complement-deactivated serum and with test serum (from LEW rats with LBN grafts) in serial dilutions. After incubation, the aliquots were taken up in glass tubes and centrifuged for 1 min. The titer was
BILLIAR, GARBEROGLIO,
AND SCHRAUT: MALTOSE ABSORPTION
77
PP-A were subject to the same sequence of histologic changes, but at a much slower rate of progression, resulting in graft fibrosis and encapsulation by the 15th to 24th day. During RESULTS the early stages of rejection the histologic Host/graft survival and graft histology. Rats changes of rejection were found to vary in with isografts survived indefinitely (> 1 year). intensity and to occur sporadically throughout These grafts showed disuse atrophy on both the allograft. macroscopic and light microscopic inspection Baseline absorption studies (Fig. I). Glucose at more than 4 months after transplantation. absorption from normal small bowel in situ Such atrophy was absent or less pronounced yielded a distinct, reproducible serum glucose in grafts examined at earlier times after trans- curve, reminiscent of that for the glucose tolplantation. Evidence of rejection was absent, erance test used in clinical practice. The MAb as expected, for isografts. The type of venous test produced a similar serum glucose curve, drainage used (PC-A vs PP-A) did not result but with a somewhat delayed and lessdistinct in any observed differences in graft histology. peak. With lactose, the peak in the serum gluFor allografis, survival times and histologic cose curve was even less distinct and was furfindings of the rejected graft differed between ther delayed; also, the curve varied markedly those with caval and those with portal venous from animal to animal and from test to test. drainage. Grafts with PC-A (N = 24) were Therefore, the lactose absorption test was disrejected after 11.8 days + 0.45 @EM). The continued. The instillation of 0.45% saline sorejection was associatedwith graft necrosisand lution alone into the small bowel was assoperitonitis resulting in the host’s death. In the ciated with a flat glucose curve. PP-A group (N = 12), longer host/graft surMaltose absorption in heterotopic isografts vival, to 22.8 days -t 1.8, was seen, the end(Fig. 2). Maltose was absorbed reproducibly point of rejection being fibrosis and encapin a given rat and in different rats for up to sulation of the graft which prevented can3 months after transplantation and was not nulation of the enterostomas and flushing of influenced significantly by the type of venous the graft. These rats were sacrificed at that drainage used (glucose level at 30 min: PCpoint. This difference between PC-A and A, 204 mg% + 8; PP-A, 213 mg% f 8; P PP-A groups is significant (P < 0.05)3 and is > 0.10). comparable to data previously obtained in our Maltose absorption in heterotopic allografts laboratory [4]. (Table 2). Allografts with PC-A demonstrated Nine rats which died (accidentally) during a reduction in their capacity for maltose abthe absorption tests were not included in the sorption on the fifth postoperative day (Fig. above-noted statistical calculations. Allografts, 3), as the glucose peak at 30 min (T30 min) with either PC-A or PP-A obtained from rats was significantly blunted in comparison to that which died on the fifth postoperative day, for isografts with PC-A (allografis: 167 mg% failed to show any histologic evidence of graft + 12; isografts: 204 mg% + 8; P < 0.02). rejection. By the sixth to eighth postoperative Functional impairment preceded histologic day, allografts with PC-A demonstrated lymchanges, which did not arise before the sixth phocytic infiltration, villous blunting, and to seventh postoperative day in rats with sporadic mucosal sloughing. These changes PC-A. In contrast, allografts with PP-A (Fig. quickly developed into complete necrosis of 4) absorbed maltose on the f&h postoperative the grafts by the 8th to 14th day. Grafts with day nearly as effectively as did isografts (T30 min for allografts: 185 mg% f 14; T30 min 3Ftests were usedto determine homogenicity of sample for isografts: 213 mg% + 8; P > 0.1). By the variance. Two-tailed Student’st testswere usedto compare parametric sample means. ninth postoperative day, the serum glucose read as the highest dilution at which agglutination persisted.
78
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984 o---o c---o 0”“Q
Maltose 4 Glucose 4 Lactose 4 045 Normd SolIne
‘\
‘pm----
100 -
0
I 15
do
/ 45
I 60
I 90
I 120
I 150
4 I 180
I 210
Time (minutes)
FIG. 1. Absorption of glucose, lactose, maltose, and 0.45% saline solution (control) from normal, in situ small bowel.
curve after maltose administration was markedly flattened for grafts with PC-A (T30 min: 137 mg% +- 11) (Fig. 3). A similar decline was not seen in the PP-A group until the 15th postoperative day (T30 min: 130 mg% + 14) (Fig. 4). In both the PC-A and PP-A groups, the peak of the serum glucose curve was delayed and was associated with progressive lowering of the glucose curve as rejection ensued. This is illustrated in Fig. 5A, which showsthe results of successive maltose absorption tests performed on an allograft with PC-A. The animal
FIG. 2. Maltose absorption test on heterotopic isods with either PP-A (N = 5) or PC-A (N = 4).
for which these curves were obtained survived for 13 days. Histologic examination at death revealed extensive lymphocytic infiltration and mucosal sloughing interspersedwith areas of seemingly intact mucosa. A blunted peak is seen in Fig. 5A on the fifth day, a delayed peak on the ninth day, and an almost flat curve on the day before death. Another PCA rat exhibited a nearly flat serum glucose curve on the ninth postoperative day (Fig. 5B), indicating extensive graft destruction and predicting the host’s death that occurred on the 1lth day. Grafts with PP-A and thus longer survival showed a lessrapid decline in maltose absorption, as shown in Fig. 5C. Flattening of the glucose curve was seenby the 15th day, indicating functional impairment because of rejection, as confirmed by histologic examination of the allograft. Immunologic studies. Anti-donor hemagglutinin titers were determined at 2- to 5-day intervals as an indicator of the intensity of the immune response of the host. Allografts with PC-A elicited a major rise in the host antidonor hemagglutinin titer between the 7th and 10th postoperative days. Such a marked increasewas not seenwith PP-A allografts until the 15th day (Fig. 6). Statistical analysis of these data (Student’s t test) revealed a signif-
BILLIAR, GARBEROGLIO, AND SCHRAUT: MALTOSE ABSORPTION
79
TABLE 2 MALTOSEABSORPTION:SERUM GLUCOSE LEVELS AT 30 MINUTES (T30 MIN) ARER MALTOSEINSTILLATION Serum glucose (mg% k SEM)
Experimental group Normal small bowel
(N = 4)
Isografts Allografts 5th postoperative day 9th postoperative day 12th postoperative day
226 f
8
PC-A (N = 4) PP-A (N = 5)
204 + 8 213f 8
PC-A (N PP-A (N PC-A (N PP-A (N PC-A (N PP-A (N
167 + 185 f 137 + 162 f 120 + 16Ok
= = = = = =
6) 6) 4) 6) 4) 5)
12 14 11 I1 14 9
icant difference between titers for PC-A and PP-A allografts for the 7th, lOth, and 15th days. DISCUSSION
A number of experimental studies have been undertaken by other investigators for determination of the functional survival as well as the morphologic integrity of small-bowel allografts. For function tests, the evaluation oftheabsorptionofglucose[5-7],glycine[7], and fatty acids [8] demonstrated impaired absorption during the course of rejection. How-
*---a ---0
200 P r
180
8z
160
1 0 E > z m
5th Fvd-op ooy emport-cqooy 12mPcd-OpDq
140 120 100
15
30
45
60
90
I20
150
180
Time (mmutes)
FIG. 3. Maltose absorption in heterotopic ahograhswith portocaval anastomosis (PC-A group).
15
50
45
60
90
120
I50
180
Time (mmutes)
FIG. 4. Maltose absorption in heterotopic allograhs with portoportal anastomosis (PP-A group).
ever, these tests, all of which involved the use of a radioactive isotope as a marker, did not allow the detection of impending or early rejection at times when morphologic changes were minor and potentially reversible. An exception appears to be the serial measurement of the absorption of “C-labeled glucose.Hardy et al. [9] found that reduced absorption paralleled the early histologic changesof rejection (edema, round cell infiltration). In contrast, Holmes et al. [6] noted that, in the presence of infection or minor histologic changesin the small-bowel allografi, the absorption of glucose was erratic. Biopsy and light-microscopic examination were thought to be better methods than measurement glucose absorption for monitoring of rejection. Recently, Cohen et al. [lo] reported a significant decreasein 14C glucose absorption from intestinal allografts in dogs, which paralleled histologic evidence of rejection. Again, however, this latter report indicates that an alteration in the pattern of glucoseabsorption is noted only when the histologic changes of rejection are already advanced. The present study indicates that the measurement of maltose absorption in the form of a glucose tolerance test can be a reliable method for monitoring of the functional integrity of small-bowel allografis which are subject to rejection. In order to be absorbed, maltose must be split into its monosaccharide, glucose, by the brush border disaccharidases;
80
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984
Time (minutes)
Time h~nutesl
I IS
I
30 45
T 60
I 90
I 120
I 150
T 190
Time (minutes)
FIG. 5. SuccessiveMAb tests on allografts. (A) Graft with PC-A surviving 13 days. (B) Graft with PPA surviving I1 days. (C) Graft with PP-A surviving 22 days.
the glucose is then absorbed actively into the portovenous circulation. In the rat and also in man, the activity of maltase is 10 times higher than that of lactase [ 11, 121;therefore, maltose intolerance does not become evident unless there is widespread mucosal impairment. Becauseof this difference in enzymatic
2
4
6
6
10
12
14
16
16
eo
FIG. 6. Rise in hemagglutinin antibody titer after smallbowel transplantation (LBN - LEW rat strain combination).
activity, we initially theorized that a lactose absorption test would be more sensitive than a maltose test and thus would be a better predictor of mucosal integrity. This expected effect was not realized in our experiments with the lactose absorption test; the serum glucose curves proved to be erratic and nondescriptive because lactose, split by lactase into glucose and galactose, had only a minor effect on the serum glucose level. In contrast, instillation of maltose led consistently to a measurable increase in serum glucose and therefore allowed it use in a function test. Both the peak and the shape of the serum glucose curve seen during the MAb test give information as to the functional competence of the graft. Seemingly predating the rejection process, or at least early in the course of rejection, the peak of the bell-shaped glucose curve is diminished and delayed. As rejection ensues and becomes clinically evident, with stoma1 edema and a hunched posture of the
BILLIAR, GARBEROGLIO, AND SCHRAUT: MALTOSE ABSORPTION
graft recipient, the curve flattens out, with a further delay of its peak. This change parallels the progressive destruction of the histologic architecture of the bowel wall, which culminates in graft necrosis. Our findings indicate that the impaired maltose absorption seen on the fifth postoperative day in allografts with PC-A predates the histologic changes (edema, lymphocytic infiltration, villous blunting) that evolve beyond the fifth postoperative day in thesegrafts. Thus, it appears that the MAb test could effectively replace mucosal biopsies, which become diagnostic of rejection at a later time. Moreover, since biopsies provide only random samples, they may not be representative of the status of the graft in the early stages of rejection. Furthermore, a biopsy specimen is not easily obtained if the graft is placed orthotopically, whereasthe MAb test allows peroral administration of the maltose solution. Although preceding histologic alterations, the impairment of maltose absorption seenon the fifth postoperative day in allografis with PC-A parallels the rise in hemagglutinin titers which also begins at that time. This relationship can be considered as additional evidence that the maltose absorption test is a sensitive indicator of early rejection. A potentially critical aspect of the maltose absorption test, and of any absorption test that relies on an intact mucosa, is the fact that a certain degree of reduction of the mucosal functional capacity and reserve, i.e., mucosal damage, must be present until a discernible and measurable impairment of absorption occurs. In this regard, a permeability test-for example, a test for leakage of a-1 antitrypsin [ 131, or the sequential measurement of the electrical resistance across the bowel wallmight prove to be still more sensitive and indicate the beginning of rejection still earlier than is possible with the MAb test. Both of these alternative approaches are being evaluated in our laboratory at present. With regard to the type of venous drainage used, we were unable to find any significant difference in maltose absorption between isografts with caval and those with portal venous
81
drainage. Theoretically, one would expect the glucose curve to rise steeply to a high peak in rats with grafts draining into the inferior vena cava, becausethe glucose reachesthe systemic circulation directly. On the other hand, in rats with portal drainage of the graft, these changes should be less pronounced, since glucose is transformed into glycogen in the liver before reaching the systemic circulation. The fact that we did not observe such effects is likely to be due to a greater, but commensurate insulin releasein responseto the rise in serum glucose in rats with grafts draining directly into the systemic circulation, thereby effecting a glucose curve equal to that seen with portal drainage of the graft. This was demonstrated by Assal et al. [ 141,who studied the effectsof portocaval shunting on the glucose tolerance test in the rat. They noted that the glucose curve after an oral glucose load was the same in shunted and in nonshunted animals, whereasthe immune-reactive insulin response was markedly elevated in shunted animals. In summary, we find that maltose absorption provides a reliable and reproducible test for monitoring of the functional integrity of small-bowel allografts, regardlessof the route of venous drainage used. In particular, in the rat model of small-bowel transplantation, in which the amount of blood that can be removed for testing is very limited, this test proved feasible and was easily performed. Becausethe maltose absorption test is a sensitive indicator of impending rejection in this animal model, it may serve as a useful function test in clinical small-bowel transplantation in the future. REFERENCES Holmes, J. T. Small-bowel transplantation: An experimental study. Ann. R. CON. Surg. Engl. 52: 165, 1973. Reznick, R. K., Craddock, Cl. N., Langer, B., et al. Structure and function of small bowel allografts in the dog. Immunosuppression with cyclosporine A. Canad. J. Surg. 25: 5 1, 1982. Rosemurgy, A. S., and Schraut, W. H. The histologic sequenceof the rejection of small-bowel allograhs in the rat. Submitted. Schraut, W. H., Rosemurgy, A. S., and Riddell, R. H. Prolongation of intestinal allograft survival
82
5.
6. 7.
8.
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984 without immunosuppressive drug therapy. J. Surg. Rex 34: 597, 1983. Cohn, W. B., Hardy, M. A., Quint, J., et al. Absorptive function in canine jejunal autografts and allografts. Surgery 65: 440, 1969. Holmes, J. T., Yeh, J., Winawer, S.J., et al. Absorption studies in canine jejunal allografts. Ann. Surg. 174: 101, 1971. Leapman, S. B., Deutsch, A. A., Grand, R. J., et al. Transplantation of fetal intestine: Survival and function in a subcutaneouslocation in adult animals. Ann. Surg. 179: 109, 1974. Stamford, W. P., and Hardy, M. A. Fatty acid absorption in jejunal autograft and allograft. Surgery
75: 496, 1974. 9. Hardy, M. A., Quint, J., and State, D. Effect of an-
tilymphocyte serum and other immunosuppressive agents on canine jejunal allografts. Ann. Surg. 171: 51, 1970.
10. Nordgren, S., Cohen, Z., Mackenzie, G. R., el al. Functional monitors of rejection in small intestinal transplantation. Amer. J. Surg. 147: 152, 1984. 11. Gray, G. M. Carbohydrate absorption and malabsorption. In L. R. Johnson (Ed.), Physiology of the Gastrointestinal Tract, Chap. 42. New York: Raven Press, 1981. P. 1063. 12. Olsen, W., Agresti, H. K., and Lorenzsonn, V. W. Intestinal disaccharidasesin diabetic rats. In R. H. Dowling and E. 0. Riecken (Eds.), Intestinal Adaptation, New York: Schattauer Verlag. P. 179. 13. Florent, C., L’Hirondel, Z., Desmazures, C., et al. Intestinal clearanceofalpha I-antitrypsin. A sensitive method for the detection of protein losing enteropathy. Gastroenterology 81: 777, 198I. 14. Assal,J. P., Levrat, R., Stauffacher,W., et al. Metabolic consequencesof portocaval shunting in the rat. Effects on glucose tolerance and serum immunoreactive insulin response.Metabolism 20: 850, 1971.
OF SURGICAL
RESEARCH
Maltose Absorption
37,75-82
(I 984)
as an indicator of Small-Intestinal
Allograft
TIMOTHY R. BILLIAR, M.D., CARLOS GARBERDGLIO, ANDWOLFGANG H. SCHRAUT,M.D.*
Rejection’
M.D.,
Department of Surgery, University of Chicago Hospitals and Clinics, 5841 South Maryland Avenue, Chicago, Illinois 60637
Presented at the Annual Meeting of the Association for Academic Surgery, Syracuse, New York, November 2-5, 1983 Maltose and lactose absorption, which require an intact brush border for breakdown and absorption as glucose, was evaluated as a function test to monitor the integrity of the small-bowel graft. Using the rat model of accessory small-bowel transplantation, absorption tests (in the form of an oral glucose tolerance test) were performed on iso- and allografts with either portal (PP-A) or caval venous drainage (PC-A). In isografts the absorption of maltose was found to be reproducible and not influenced by the type of venous drainage. This was not the case with the use of lactose which thus was not studied further. Allografts with PC-A demonstrated a reduction in their capacity for maltose absorption on the fifth postoperative day, as the glucose peak at 30 min (T30) was significantly blunted in comparison to that for isografts with PC-A (167 mg% 2 12 vs 204 mg% 5 8). Functional impairment preceded histologic changeswhich did not arise before the sixth-to-seventh postoperative day in rats with PC-A. Allografis with PP-A absorbed maltose on the fifth postoperative day nearly as effectively as did isografts (T30 min: 185 mg% ? 14 vs 213 mg% + 8). By the ninth postoperative day, the serum glucose curve after maltose administration was flattened for grafts with PC-A (T30 min: I37 mg% f 1I) which were rejected acutely (host’s death) after 11.8 days + 0.45. A similar impairment of maltose absorption was not seen in the PP-A group (chronic graft rejection after 22.8 days + 1.8) until the 15th postoperative day. These results suggestthat maltose absorption becomes impaired before the clinical appearance of rejection and its histologic alterations. Tbe maltose absorption test effectively predicts rejection and monitors the function of the intestinal graft in this animal model.
elusively. Furthermore, this method of monitoring requires a cutaneous enterostoma, The expectant monitoring of small intessince orthotopically placed grails are relatively tinal allografts for rejection presentsa problem inaccessible for frequent biopsy. In addition, not seen with other types of organ transplana small-bowel graft that appears morphologtation becausea simple, reproducible function ically to be intact does not necessarily have test is unavailable-a test that would reflect functional integrity. functional and thus morphologic integrity of We postulate that intestinal absorption of the allog&. At present, histologic evaluation the disaccharides, maltose and lactose, may of serial mucosal biopsy specimensis the most provide a sensitive function test. This test recommon method of following the rejection quires the enzymes (disaccharidases) of the process. However, at least in canine and rat mucosal brush border to be intact for the intestinal allografts, the histologic alterations breakdown of maltose and lactose into glucose that occur during rejection are known to be and fructose and for the active absorption of variable in both distribution and intensity these monosaccharides. In this study on a rat [l-3]. This indicates that random biopsies may model of accessory small-bowel transplantafail to reflect the status of the allograft contion, we measured the absorption of maltose ’ Supported by the National Institutes of Arthritis, Me- and lactose in the form of a glucose tolerance test. We sought to determine whether this test tabolism and Digestive Diseases(Grant AM27332). ’ To whom requests for reprints should be addressed. can be used as a method for predicting and INTRODUCTION
75
0022-4804/84 $1.50 Copyright Q 1984 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984
16
monitoring the rejection of small-bowel allografts as it becomes evident from histologic and serum-immunologic studies. MATERIALS
AND METHODS
We used rats of the inbred Lewis (LEW) and Lewis X Brown-Norway F- 1 hybrid (LBN) strains, weighing between 180 and 250 g. Absorption studies were carried out in the normal (nontransplanted, in situ) small bowel, in insografts, and in allogrtis (Table 1). Operative procedures. For absorption studies on the nontransplanted small bowel, LEW rats were used. Under chloral hydrate [3 mg/gm body wt (SW)] anesthesia, a Silastic catheter was placed via an enterostomy into the proximal jejunum of the fasted animal. The bowel was irrigated with 0.45% saline solution and then was occluded proximally (duodenum) and distally (ileocecal junction). This was followed by instillation of the test solution. After each absorption test, the enterostomy site was closed and the animal allowed to recuperate. Three tests per animal were undertaken at I- to 2-week intervals. The entire small bowel was transplanted as an accessory graft according to a procedure described previously [4]. Isografts and allografts, the latter with the LEB - LEW strain TABLE I Experimental group
No. of rats
Normal, nontransplanted small bowel
4
Small-bowel isografts with Portocaval anastomosis (PC-A) Portoportal anastomosis (PP-A)
4 5
Total Small-bowel allografts with Portocaval anastomosis (PC-A) Portoportal anastomosis (PP-A) Total
(9) 24” 12” (36)’
a Absorption studies were carried out on only six rats in each group. ‘Excludes nine rats that died during the absorption tests.
combination, were transplanted with provision made for either portal (portoportal anastomosis, PP-A) or systemic (portocaval) anastomosis, PC-A) venous drainage. Absorption tests.Baseline absorption studies were performed with normal saline, glucose, lactose, and maltose in the small bowel of normal rats (N = 4). Glucose (1 mg/g BW), lactose (1 mg/g BW), or maltose (0.5 mg/g BW) was dissolved in 2.5 ml of 0.45% saline solution and instilled into the bowel. Serum glucose levels were measured with a Beckman glucoseanalyzer at 5- to 15-min intervals until the glucose level returned to the previously recorded baseline. Since the initial measurements showed that the absorption of lactose was highly variable from test to test, we limited the absorption studies in rats with iso- and allografts to the use of maltose as substrate. In rats with isografts the maltose absorption (MAb) test was performed repeatedly (three to four times) on random days from the seventh postoperative day up to 3 months postoperatively. In 12 of the 40 rats with allografts, the test was undertaken successively on the 5th, 9th, 12th, and 15th postoperative days, unless graft rejection prevented the next test. On the day before and the day of the MAb test, the graft was irrigated clear with 5- 10 cc of normal saline. The maltose solution was infused into the proximal stoma, and both the proximal and distal stoma were then occluded with balloon catheters. Histologic examination. Upon death or sacrifice of the rats, all grafts were fixed in formalin, and sections were prepared (H&E stain) for light microscopy. Hemagglutinin antibody assay. In order to have a measure of the host’s immune reaction to the small-bowel allograft, we measuredantidonor hemagglutinin titers. Washed LBN erythrocytes were incubated for 45 min with normal (LEW) complement-deactivated serum and with test serum (from LEW rats with LBN grafts) in serial dilutions. After incubation, the aliquots were taken up in glass tubes and centrifuged for 1 min. The titer was
BILLIAR, GARBEROGLIO,
AND SCHRAUT: MALTOSE ABSORPTION
77
PP-A were subject to the same sequence of histologic changes, but at a much slower rate of progression, resulting in graft fibrosis and encapsulation by the 15th to 24th day. During RESULTS the early stages of rejection the histologic Host/graft survival and graft histology. Rats changes of rejection were found to vary in with isografts survived indefinitely (> 1 year). intensity and to occur sporadically throughout These grafts showed disuse atrophy on both the allograft. macroscopic and light microscopic inspection Baseline absorption studies (Fig. I). Glucose at more than 4 months after transplantation. absorption from normal small bowel in situ Such atrophy was absent or less pronounced yielded a distinct, reproducible serum glucose in grafts examined at earlier times after trans- curve, reminiscent of that for the glucose tolplantation. Evidence of rejection was absent, erance test used in clinical practice. The MAb as expected, for isografts. The type of venous test produced a similar serum glucose curve, drainage used (PC-A vs PP-A) did not result but with a somewhat delayed and lessdistinct in any observed differences in graft histology. peak. With lactose, the peak in the serum gluFor allografis, survival times and histologic cose curve was even less distinct and was furfindings of the rejected graft differed between ther delayed; also, the curve varied markedly those with caval and those with portal venous from animal to animal and from test to test. drainage. Grafts with PC-A (N = 24) were Therefore, the lactose absorption test was disrejected after 11.8 days + 0.45 @EM). The continued. The instillation of 0.45% saline sorejection was associatedwith graft necrosisand lution alone into the small bowel was assoperitonitis resulting in the host’s death. In the ciated with a flat glucose curve. PP-A group (N = 12), longer host/graft surMaltose absorption in heterotopic isografts vival, to 22.8 days -t 1.8, was seen, the end(Fig. 2). Maltose was absorbed reproducibly point of rejection being fibrosis and encapin a given rat and in different rats for up to sulation of the graft which prevented can3 months after transplantation and was not nulation of the enterostomas and flushing of influenced significantly by the type of venous the graft. These rats were sacrificed at that drainage used (glucose level at 30 min: PCpoint. This difference between PC-A and A, 204 mg% + 8; PP-A, 213 mg% f 8; P PP-A groups is significant (P < 0.05)3 and is > 0.10). comparable to data previously obtained in our Maltose absorption in heterotopic allografts laboratory [4]. (Table 2). Allografts with PC-A demonstrated Nine rats which died (accidentally) during a reduction in their capacity for maltose abthe absorption tests were not included in the sorption on the fifth postoperative day (Fig. above-noted statistical calculations. Allografts, 3), as the glucose peak at 30 min (T30 min) with either PC-A or PP-A obtained from rats was significantly blunted in comparison to that which died on the fifth postoperative day, for isografts with PC-A (allografis: 167 mg% failed to show any histologic evidence of graft + 12; isografts: 204 mg% + 8; P < 0.02). rejection. By the sixth to eighth postoperative Functional impairment preceded histologic day, allografts with PC-A demonstrated lymchanges, which did not arise before the sixth phocytic infiltration, villous blunting, and to seventh postoperative day in rats with sporadic mucosal sloughing. These changes PC-A. In contrast, allografts with PP-A (Fig. quickly developed into complete necrosis of 4) absorbed maltose on the f&h postoperative the grafts by the 8th to 14th day. Grafts with day nearly as effectively as did isografts (T30 min for allografts: 185 mg% f 14; T30 min 3Ftests were usedto determine homogenicity of sample for isografts: 213 mg% + 8; P > 0.1). By the variance. Two-tailed Student’st testswere usedto compare parametric sample means. ninth postoperative day, the serum glucose read as the highest dilution at which agglutination persisted.
78
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984 o---o c---o 0”“Q
Maltose 4 Glucose 4 Lactose 4 045 Normd SolIne
‘\
‘pm----
100 -
0
I 15
do
/ 45
I 60
I 90
I 120
I 150
4 I 180
I 210
Time (minutes)
FIG. 1. Absorption of glucose, lactose, maltose, and 0.45% saline solution (control) from normal, in situ small bowel.
curve after maltose administration was markedly flattened for grafts with PC-A (T30 min: 137 mg% +- 11) (Fig. 3). A similar decline was not seen in the PP-A group until the 15th postoperative day (T30 min: 130 mg% + 14) (Fig. 4). In both the PC-A and PP-A groups, the peak of the serum glucose curve was delayed and was associated with progressive lowering of the glucose curve as rejection ensued. This is illustrated in Fig. 5A, which showsthe results of successive maltose absorption tests performed on an allograft with PC-A. The animal
FIG. 2. Maltose absorption test on heterotopic isods with either PP-A (N = 5) or PC-A (N = 4).
for which these curves were obtained survived for 13 days. Histologic examination at death revealed extensive lymphocytic infiltration and mucosal sloughing interspersedwith areas of seemingly intact mucosa. A blunted peak is seen in Fig. 5A on the fifth day, a delayed peak on the ninth day, and an almost flat curve on the day before death. Another PCA rat exhibited a nearly flat serum glucose curve on the ninth postoperative day (Fig. 5B), indicating extensive graft destruction and predicting the host’s death that occurred on the 1lth day. Grafts with PP-A and thus longer survival showed a lessrapid decline in maltose absorption, as shown in Fig. 5C. Flattening of the glucose curve was seenby the 15th day, indicating functional impairment because of rejection, as confirmed by histologic examination of the allograft. Immunologic studies. Anti-donor hemagglutinin titers were determined at 2- to 5-day intervals as an indicator of the intensity of the immune response of the host. Allografts with PC-A elicited a major rise in the host antidonor hemagglutinin titer between the 7th and 10th postoperative days. Such a marked increasewas not seenwith PP-A allografts until the 15th day (Fig. 6). Statistical analysis of these data (Student’s t test) revealed a signif-
BILLIAR, GARBEROGLIO, AND SCHRAUT: MALTOSE ABSORPTION
79
TABLE 2 MALTOSEABSORPTION:SERUM GLUCOSE LEVELS AT 30 MINUTES (T30 MIN) ARER MALTOSEINSTILLATION Serum glucose (mg% k SEM)
Experimental group Normal small bowel
(N = 4)
Isografts Allografts 5th postoperative day 9th postoperative day 12th postoperative day
226 f
8
PC-A (N = 4) PP-A (N = 5)
204 + 8 213f 8
PC-A (N PP-A (N PC-A (N PP-A (N PC-A (N PP-A (N
167 + 185 f 137 + 162 f 120 + 16Ok
= = = = = =
6) 6) 4) 6) 4) 5)
12 14 11 I1 14 9
icant difference between titers for PC-A and PP-A allografts for the 7th, lOth, and 15th days. DISCUSSION
A number of experimental studies have been undertaken by other investigators for determination of the functional survival as well as the morphologic integrity of small-bowel allografts. For function tests, the evaluation oftheabsorptionofglucose[5-7],glycine[7], and fatty acids [8] demonstrated impaired absorption during the course of rejection. How-
*---a ---0
200 P r
180
8z
160
1 0 E > z m
5th Fvd-op ooy emport-cqooy 12mPcd-OpDq
140 120 100
15
30
45
60
90
I20
150
180
Time (mmutes)
FIG. 3. Maltose absorption in heterotopic ahograhswith portocaval anastomosis (PC-A group).
15
50
45
60
90
120
I50
180
Time (mmutes)
FIG. 4. Maltose absorption in heterotopic allograhs with portoportal anastomosis (PP-A group).
ever, these tests, all of which involved the use of a radioactive isotope as a marker, did not allow the detection of impending or early rejection at times when morphologic changes were minor and potentially reversible. An exception appears to be the serial measurement of the absorption of “C-labeled glucose.Hardy et al. [9] found that reduced absorption paralleled the early histologic changesof rejection (edema, round cell infiltration). In contrast, Holmes et al. [6] noted that, in the presence of infection or minor histologic changesin the small-bowel allografi, the absorption of glucose was erratic. Biopsy and light-microscopic examination were thought to be better methods than measurement glucose absorption for monitoring of rejection. Recently, Cohen et al. [lo] reported a significant decreasein 14C glucose absorption from intestinal allografts in dogs, which paralleled histologic evidence of rejection. Again, however, this latter report indicates that an alteration in the pattern of glucoseabsorption is noted only when the histologic changes of rejection are already advanced. The present study indicates that the measurement of maltose absorption in the form of a glucose tolerance test can be a reliable method for monitoring of the functional integrity of small-bowel allografis which are subject to rejection. In order to be absorbed, maltose must be split into its monosaccharide, glucose, by the brush border disaccharidases;
80
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984
Time (minutes)
Time h~nutesl
I IS
I
30 45
T 60
I 90
I 120
I 150
T 190
Time (minutes)
FIG. 5. SuccessiveMAb tests on allografts. (A) Graft with PC-A surviving 13 days. (B) Graft with PPA surviving I1 days. (C) Graft with PP-A surviving 22 days.
the glucose is then absorbed actively into the portovenous circulation. In the rat and also in man, the activity of maltase is 10 times higher than that of lactase [ 11, 121;therefore, maltose intolerance does not become evident unless there is widespread mucosal impairment. Becauseof this difference in enzymatic
2
4
6
6
10
12
14
16
16
eo
FIG. 6. Rise in hemagglutinin antibody titer after smallbowel transplantation (LBN - LEW rat strain combination).
activity, we initially theorized that a lactose absorption test would be more sensitive than a maltose test and thus would be a better predictor of mucosal integrity. This expected effect was not realized in our experiments with the lactose absorption test; the serum glucose curves proved to be erratic and nondescriptive because lactose, split by lactase into glucose and galactose, had only a minor effect on the serum glucose level. In contrast, instillation of maltose led consistently to a measurable increase in serum glucose and therefore allowed it use in a function test. Both the peak and the shape of the serum glucose curve seen during the MAb test give information as to the functional competence of the graft. Seemingly predating the rejection process, or at least early in the course of rejection, the peak of the bell-shaped glucose curve is diminished and delayed. As rejection ensues and becomes clinically evident, with stoma1 edema and a hunched posture of the
BILLIAR, GARBEROGLIO, AND SCHRAUT: MALTOSE ABSORPTION
graft recipient, the curve flattens out, with a further delay of its peak. This change parallels the progressive destruction of the histologic architecture of the bowel wall, which culminates in graft necrosis. Our findings indicate that the impaired maltose absorption seen on the fifth postoperative day in allografts with PC-A predates the histologic changes (edema, lymphocytic infiltration, villous blunting) that evolve beyond the fifth postoperative day in thesegrafts. Thus, it appears that the MAb test could effectively replace mucosal biopsies, which become diagnostic of rejection at a later time. Moreover, since biopsies provide only random samples, they may not be representative of the status of the graft in the early stages of rejection. Furthermore, a biopsy specimen is not easily obtained if the graft is placed orthotopically, whereasthe MAb test allows peroral administration of the maltose solution. Although preceding histologic alterations, the impairment of maltose absorption seenon the fifth postoperative day in allografis with PC-A parallels the rise in hemagglutinin titers which also begins at that time. This relationship can be considered as additional evidence that the maltose absorption test is a sensitive indicator of early rejection. A potentially critical aspect of the maltose absorption test, and of any absorption test that relies on an intact mucosa, is the fact that a certain degree of reduction of the mucosal functional capacity and reserve, i.e., mucosal damage, must be present until a discernible and measurable impairment of absorption occurs. In this regard, a permeability test-for example, a test for leakage of a-1 antitrypsin [ 131, or the sequential measurement of the electrical resistance across the bowel wallmight prove to be still more sensitive and indicate the beginning of rejection still earlier than is possible with the MAb test. Both of these alternative approaches are being evaluated in our laboratory at present. With regard to the type of venous drainage used, we were unable to find any significant difference in maltose absorption between isografts with caval and those with portal venous
81
drainage. Theoretically, one would expect the glucose curve to rise steeply to a high peak in rats with grafts draining into the inferior vena cava, becausethe glucose reachesthe systemic circulation directly. On the other hand, in rats with portal drainage of the graft, these changes should be less pronounced, since glucose is transformed into glycogen in the liver before reaching the systemic circulation. The fact that we did not observe such effects is likely to be due to a greater, but commensurate insulin releasein responseto the rise in serum glucose in rats with grafts draining directly into the systemic circulation, thereby effecting a glucose curve equal to that seen with portal drainage of the graft. This was demonstrated by Assal et al. [ 141,who studied the effectsof portocaval shunting on the glucose tolerance test in the rat. They noted that the glucose curve after an oral glucose load was the same in shunted and in nonshunted animals, whereasthe immune-reactive insulin response was markedly elevated in shunted animals. In summary, we find that maltose absorption provides a reliable and reproducible test for monitoring of the functional integrity of small-bowel allografts, regardlessof the route of venous drainage used. In particular, in the rat model of small-bowel transplantation, in which the amount of blood that can be removed for testing is very limited, this test proved feasible and was easily performed. Becausethe maltose absorption test is a sensitive indicator of impending rejection in this animal model, it may serve as a useful function test in clinical small-bowel transplantation in the future. REFERENCES Holmes, J. T. Small-bowel transplantation: An experimental study. Ann. R. CON. Surg. Engl. 52: 165, 1973. Reznick, R. K., Craddock, Cl. N., Langer, B., et al. Structure and function of small bowel allografts in the dog. Immunosuppression with cyclosporine A. Canad. J. Surg. 25: 5 1, 1982. Rosemurgy, A. S., and Schraut, W. H. The histologic sequenceof the rejection of small-bowel allograhs in the rat. Submitted. Schraut, W. H., Rosemurgy, A. S., and Riddell, R. H. Prolongation of intestinal allograft survival
82
5.
6. 7.
8.
JOURNAL OF SURGICAL RESEARCH: VOL. 37, NO. 1, JULY 1984 without immunosuppressive drug therapy. J. Surg. Rex 34: 597, 1983. Cohn, W. B., Hardy, M. A., Quint, J., et al. Absorptive function in canine jejunal autografts and allografts. Surgery 65: 440, 1969. Holmes, J. T., Yeh, J., Winawer, S.J., et al. Absorption studies in canine jejunal allografts. Ann. Surg. 174: 101, 1971. Leapman, S. B., Deutsch, A. A., Grand, R. J., et al. Transplantation of fetal intestine: Survival and function in a subcutaneouslocation in adult animals. Ann. Surg. 179: 109, 1974. Stamford, W. P., and Hardy, M. A. Fatty acid absorption in jejunal autograft and allograft. Surgery
75: 496, 1974. 9. Hardy, M. A., Quint, J., and State, D. Effect of an-
tilymphocyte serum and other immunosuppressive agents on canine jejunal allografts. Ann. Surg. 171: 51, 1970.
10. Nordgren, S., Cohen, Z., Mackenzie, G. R., el al. Functional monitors of rejection in small intestinal transplantation. Amer. J. Surg. 147: 152, 1984. 11. Gray, G. M. Carbohydrate absorption and malabsorption. In L. R. Johnson (Ed.), Physiology of the Gastrointestinal Tract, Chap. 42. New York: Raven Press, 1981. P. 1063. 12. Olsen, W., Agresti, H. K., and Lorenzsonn, V. W. Intestinal disaccharidasesin diabetic rats. In R. H. Dowling and E. 0. Riecken (Eds.), Intestinal Adaptation, New York: Schattauer Verlag. P. 179. 13. Florent, C., L’Hirondel, Z., Desmazures, C., et al. Intestinal clearanceofalpha I-antitrypsin. A sensitive method for the detection of protein losing enteropathy. Gastroenterology 81: 777, 198I. 14. Assal,J. P., Levrat, R., Stauffacher,W., et al. Metabolic consequencesof portocaval shunting in the rat. Effects on glucose tolerance and serum immunoreactive insulin response.Metabolism 20: 850, 1971.