- Email: [email protected]
Lymphoid depletion in ileal loops: Mechanism and clinical implications
Lymphoid Depletion in Neal Loops: Mechanism and Clinical Implications By David Tapper
and Judah Folkman
N
1926 Seiffert first described the construction of ileal loop conduits for urinary diversion.’ Bricker in the 1950s popularized the use of the ileal loop as a means of supravesical urinary diversion following exenteration for pelvic malignancy in adults.* Bill and Nash as well as others suggested its usefulness in children with nonmalignant disease of the urinary tract.3l4 Many ileal loop conduits have been constructed in children since 1952. There are numerous early reports and several long-term studies of the results of this surgical procedure in children. 5m” In almost all reports there is a 10Y;P20”,/, incidence of renal deterioration after the construction of the loop. It is not clear why some patients with normal renal function deteriorate after creation of the loop. Although a stoma1 or uretero-ileal stenosis is one obvious cause, in most of these patients such derangement cannot be demonstrated. In addition, only lS’{ of patients with already diminished renal function improve after diversion; the remainder are unchanged or deteriorate further.12 Three patterns of complications stand out: recurrence of urinary tract infection, continued deterioration of renal function, and late stenosis or stricture of the midportion of the ileal loop conduit. We have had extensive experience with the creation of ileal loops at The Children’s Hospital Medical Center. Our policy has been that those children who do not have a neurogenic bladder be considered for urinary dediversion.13 We had the opportunity to study intestinal segments exposed to urine and to compare them to intestinal segments no longer exposed to urine because of diversion. We suggest that the recurrent urinary tract infections, the deteriorating renal function, and the midloop stenosis are related to the same underlying problem, namely a severe depletion of lymphoid elements in the intestine. Our studies indicate that it is urine which causes the lymphoid depletion. We have tested this hypothesis clinically and experimentally. I
METHOD
Clinical Studies. Since 1952. 162 ileal loop diversions have been performed at Children’s Hospital Medical Center. Sixteen of the 162 patients underwent dediversion and reestablishment of urinary continuity, and represent the basis for this study. Histories have been previously reported for I1 of the 16.13 The 16 cases were observed at three stages in their clinical course (Fig. IA).
From the Departments of Surgey. Children’s Hospital Medical Center and The Harvard Medical School, Boston. Mass. and Universily of Cali’omia, San Francisco Medical Center, San Francisco. Calif Presented beJore the 7th Annual Meeting of’ the American Pediatric Surgical Association. Boca Raion. Fla.s April 29-Ma)l I, 1976. Address for reprinr requests: Judah Folkman. M.D.. The Children’s Hospital Medical Center, 300 Longwood Avenue, Boston. Mass. 021 IS. w 1976 bv Gntne & Stratron. Inc. Journal of Pediatric Surgery, Vol. 11, No. 5 (October), 1976
871
TAPPER
872
AND FOLKMAN
STAGE I
STAGE II
STAGE III
STAGE I
STAGE II
STAGE III
A
B Fig. 1. (A) Representation of the three operations performed in children undergoing creation of Real loop. Stage I: formation of standard Real loop; stage II: dediversion and reestablishment of urinary continuity; and stage Ill: excision of defunctionalized intestine. (8) Animal model. Stage I: creation of right uretero-ileal anastomosis with ileal loop; stage II: nephroureterectomy with excision of proximal portion of loop; ond stage Ill: excision of defunctionalized intestine.
Stage I: Creation of the ileal loop. The ureter or pelvic calyx was anastomosed ileum with urine draining through the intestine and exiting through a stoma abdominal wall. Stage II: Reestablishment of urinary continuity. Six reconstructions were
to the terminal on the anterior carried
out
by
ureteroureteral anastomosis. One child had an ileocecal-cystoplasty, one had an ileocystoplasty, one had a renal autotransplantation, one had the loop implanted into the bladder, and one had the loop anastomosed to the distal ureteral stump. Stage III: Excision of dejiinctionalized intestinal segment. At least 1 yr after diversion the defunctionalized ileum was removed, This was done only if it was certain that diversion would not again be necessary, and if the loop was not utilized in the reconstruction.
Experimental
Studies
Eleven mongrel dogs weighing 15 kg were similarly staged (Fig. 1B). A combination of intravenous diabutal and open drop ether was used for anesthesia. segment of terminal ileum approxiStage I: Creation of the ileal loop. An appropriate mately 8 cm from the ileocecal valve was selected for the conduit. The blind end of the loop was closed. The right ureter was anastomosed to the back wall near the blind end of the ileal conduit, using mucosa-to-mucosa interrupted sutures of 5-O chromic catgut. The loop was allowed to lie confortably in front of, or behind the cecum. The open end was brought out through a previously constructed stoma site and matured primarily. Intestinal continuity was reestablished.
LYMPHOID
073
DEPLETION
Two weeks after the first operation the dogs were reanesthetized, the loop catheterized, urine residual measured, and serum creatinine determined. After an interval of 4-12 wk Stage II: Partial excision of the loop and nephroureterectomv. (mean 8 wk) a second laparotomy was performed. At this operation the uretero-ileal anastomosis was inspected. A right nephroureterectomy was performed which included approximately 2576 of the proximal portion of the ileal loop and contained the uretero-ileal anastomosis. The blind end of the loop was closed. Histologic sections were prepared from the entire ileum and stained with hematoxylin and eosin. Stage III: Excision of remainder of ileal loop. After 4-8 wk (mean 5 wk) a third laparotomy was performed. At this operation the remaining segment of the loop was excised. Four animals served as controls. In two dogs intestinal loops were created from a similar portion of the terminal ileum. However, the loop was essentially a defunctionalized portion of intestine without any urine flow. In two animals two loops were created; one loop had a uretero-ileal anastomosis with urine flowing through it: the other loop was defunctionalized intestine.
RESULTS
Clinical Study
Sixteen patients underwent dediversion from their iieal loops and reconstruction of the urinary system. Unfortunately, sections of the intestine were not available at each stage of the procedure in every patient. In addition several patients underwent interval procedures to revise the ileal loop. However, results were identical in all patients studied. Stage I. Figure 2 is a representation of the changes seen in the ileal loop intestinal segments prior to urine exposure, after urine exposure, and after dediversion. Figure 2A depicts the normal Peyer’s patches as seen in the terminal ileum of human intestine. The Peyer’s patches are made up of mature lymphoid elements coalescing to form follicles with healthy germinal centers. There also is a layer of mature lymphocytes beneath the intestinal villi (Fig. 2B). Stage II. The effect of urine exposure on the lymphoid elements in the intestine is evident in Fig. 2C. There is a marked depletion of cellular elements. The follicles are smaller and depleted of mature, active cells. The layer of lymphoid elements usually noted beneath the villi contains only scanty lymphoid cells (Fig. 2D). In the histologic sections there was no evidence of lymphoid dilatation to suggest leakage of lymphocytes. Stage III. The ileal segment remains in situ and defunctionalized for at least 1 yr following dediversion. During this absence of urine, the follicles have regenerated (Fig. 2E). The lymphoid cells regain their plump, healthy appearance. The layer above the lamina propria has been repopulated with lymphoid elements (Fig. 2F). Eosinophils are present in the histologic sections which suggest a reparative process. Experimental Study
Ten of the eleven dogs survived. One dog died of distemper. Cannulation of the loop at 2 wk after the first laparotomy showed prompt urine flow with only a 5-10 cc residual in nine out of ten dogs. One animal had slight stoma1 narrowing but catheterization was easy and the loop contained only 10 cc residual. Dog #ll in which no urine was obtained from the loop will be discussed later.
TAPPER
874
I
STAGE I
STAGE II
STAGE
AND
FOLKMAN
III
ILEUM EXPOSURE
NO URINE
TO
NO URINE
URINE 3
Muscle
:
z c,rvm
Fig. 2. Representation of effect of urine on the lymphoid elements in the ileal loops of children undergoing staged dediversion. (A) Confluence of lymphoid follicles to form peyer’s patches with active germinal centers prior to urine exposure. (B) layer of healthy lymphocytes beneath the intestinal villi prior to urine exposure. (C) Marked depletion of lymphoid elements after exposure to urine. (D) Loss of lymphocytes from beneath the villi with urine exposure. (E) Reappearance of lymphoid follicles with removal of urine. (F) Regeneration of lymphocytes beneath the villi.
Stage I. At the first laparotomy a biopsy was taken from the proximal portion of the loop. This served as control intestine, “before exposure to urine.” The Peyer’s patches in the canine terminal ileum are larger and more numerous than those in the human intestine; they also appear more confluent (Fig. 3A). Most contain large follicles with large, active germinal centers. The lymphocytes making up the Peyer’s patches are plump, mature, and healthy. Just above the lamina propria is a second layer of lymphocytes. This layer is approximately 668 cells thick and is beneath the villi (Fig. 3B). At the villous tips there are more plump, mature lymphocytes. Stage II. After approximately 8 wk, during which time the ileal loop was continuously exposed to urine, the second laparotomy was performed. Histologic sections were made on the entire portion of the resected ileum containing the uretero-ileal anastomosis. The results were similar in all the animals. Figure 3C is representative of the effect of urine on the intestine especially on the lymphoid elements. There is marked depletion of all lymphoid elements espe-
LYMPHOID
STAGE
875
DEPLETION
STAGE III
STAGE II
I
ILEUM NO URINE
Fig. 3. loops. (A)
RE TO
Representation of the effect of urine on the lymphoid elements in the canine ileal Confluent Peyer’s patches as seen in normal canine intestine prior to urine expo-
sure. (B) lymphocytes beneath the intestinal villi prior to urine exposure. (C) Severe lymphoid depletion and toss of follicle architecture with exposure to urine. (D) Depletion of lymphocytes from beneath the villi. (E) Reappearance of lymphoid follicles with active germinal centers with the removal of urine exposure. (F) lymphocytes beginning to reappear beneath intestinal villi as urine exposure is removed.
cially within the follicles. Most of them have been destroyed. The remaining cells are no longer mature and healthy; there are only fine reticular lymphocytes. In the histologic sections there was no evidence of lymphoid dilatation. The lymphoid cell layer just beneath the villi has disappeared and only a few scattered lymphocytes remain (Fig. 3D). Stage ZIZ. After a period of approximately 5 wk, during which time no urine flowed through the loop, the third laparotomy was performed. Figure 3E is a representative section and shows the cellular repopulation taking place. Lymphoid cells are reaccumulating beneath the villi. There are small new follicles forming (Fig. 3F). A moderate number of eosinophils related to the recovery process are also present in the histologic preparations. Figure 4 is from Dog #l which had the longest recovery period (8 wk). Restoration of follicles continues to progress even 8 wk after removal of urine from the loop. More lymphocytes are forming along the border of the crypts and as if by a “homing signal” they are coalescing to form follicles. At the second laparotomy we noticed hypertrophy of the lymph nodes confined to the mesentery of the ileal loop segment (Fig. 5). There was no evidence
876
TAPPER
AND FOLKMAN
Fig. 4. The lymphocytes beneath the villi appear to “home in” and to reform follicles once the urine exposure is removed. This appears to be time dependent.
lymphoid
of a generalized lymphadenopathy. We did not see hypertrophy of mesenteric lymph nodes in the control animals without urine exposure. Hematoxin- eosinstained sections showed reactive hyperplasia of the lymph nodes without any evidence of granuloma formation. When the final portions of ileum were removed at the third laparotomy, the nodes in the mesentery had regressed and were no longer easily visible. The intestine from the control animals showed no evidence of atrophy, ischemia or depletion of lymphoid elements. The number of mature lymphoid elements remained unchanged. Dog #11 which did not make urine through the loop was staged similarly STAGE
I
I
STAGE
II
When urine expoFig. 5. sure is present, there is an associated hypertrophy of the lymph nodes in the mesentery of the ileal loop. The hypertrophic nodes shrink when urine exposure is removed.
LYMPHOID
DEPLETION
a77
Fig. 6. Kidney from dog which did not make x11 urine. Note the large area of segmental infarction in the renal parenchyma.
to the other animals. At the second laparotomy there was no evidence of ureteral obstruction or hydronephrosis. On the contrary the kidney was markedly smaller on the right side. The uretero-ileal anastomosis was patent and only a small amount of methylene blue was noted to come down the ureter. Figure 6 is a section of the involved kidney and shows segmental ischemia. The damage was probably secondary to segmental renal artery stenosis or occlusion which was not detected at the first laparotomy. No dissection was done in the retroperitoneum at the first stage but the possibility of an iatrogenic injury cannot be ruled out. The pathology of the diverted intestinal loop was similar to the control animals without urine exposure. The presence of ureteral epithelium does not alter morphology in the loop. Serum creatinine values were normal. There was no evidence of renal failure in any of the experimental animals. No systemic infections occurred in any of the dogs. Urinalysis at the time of loop catheterizations showed epithelial debris, and a small number of white cells. No colony counts were done. DlSCUSSlON
These studies demonstrate that exposure of intestinal segments to urine causes marked lymphoid depletion in the segments. When the urine exposure is interrupted as by dediversion in patients or nephroureterectomy in dogs, the
878
TAPPER
AND
FOLKMAN
lymphocytes return. These studies give additional support to the idea advanced by others that a lymphocyte suppressive factor exists in urine.r4 If such a factor exists in urine and causes lymphocyte depletion, this may help explain the occurrence of late midloop strictures, recurrent infections, and continued renal deterioration. Peyer’s patches have tremendous regenerative capacity as demonstrated by the return of lymphoid elements after the urine was removed. Jacobsen showed that if 1 Peyer’s patch in a mouse was exteriorized and protected with a lead shield, he could give normally lethal doses of whole body irradiation without fatality.15 The mouse repopulated his entire lymphoid system from 1 protected Peyer’s patch. The continued presence of urine bathing the intestinal mucosa appears to locally inhibit regeneration of the Peyer’s patches. In addition the loss of the lymphocyte barrier in the intestinal segment may predispose the lymph nodes in the ileal loop mesentery to hypertrophy. These hypertrophied nodules may be responsible for late midloop strictures. Hardy has reported 17 cases of late midloop strictures. These occurred in patients several years after construction of their ileal 10op.l~ Loopograms on these patients showed midloop narrowing and stenosis with proximal dilatation. There was no evidence of vascular compromise. All of these loops were removed because of deteriorating renal function. The mesentery of these loops contained hypertrophied lymph nodes which seemed to encroach and narrow the intestinal lumen. This finding of hypertrophied mesenteric nodes has been confirmed by others.” This reactive hyperplasia of the lymph nodes in the heal mesentery may be due to lymphoid depletion in the intestinal segment. The mesenteric nodes would then act as a second line of defense against organisms which invaded the ileal mucosa. The ileal loop segment with its hypertrophied nodes looks similar to intestine involved in regional enteritis. The small intestine is normally resistant to infection. However, the loss of the villous architecture and the depletion of lymphocytes from the Peyer’s patches make the intestinal segment little more than a conduit. Peyer’s patches are immunocompetent lymphoid organs which participate in intestinal immune responses. I8 The depletion of these lymphoid cells coupled with the refluxing nature of the ileal conduit may lead to recurrent infection and continued kidney damage. Current treatment focuses on the upper urinary tracts. Systemic antibiotics are given to decrease bacterial flora in the urine and to protect the kidney from further damage. It may be fruitful for pediatric surgeons and urologists to reexamine conventional treatment for ileal loop infections and for midloop stenosis. Local therapy in the form of nonabsorbable intestinal antibiotics may prevent further damage to renal parenchyma. The conventional treatment for midloop stenosis with proximal-loop dilatation is resection of the loop and construction of a new ileal conduit. However, it may be possible to use a combination of steroid enemas and nonabsorbable antibiotics to shrink hypertrophied mesenteric nodes and relieve obstruction before a permanent stricture develops. This might prevent an additional operation to fashion a new ileal loop.
LYMPHOlD
DEPLETION
a79
SUMMARY
Recent long-term follow-up studies of children undergoing ileal loop urinary diversion have shown a number of complications arising in these patients: recurrent urinary tract infections, deteriorating renal function, and late midloop stricture. We suggest that all three may be related to the same underlying problem, namely, a severe depletion of lymphoid elements in the intestine. In both clinical and animal studies urine exposure to intestinal segments resulted in a marked depletion of lymphoid elements from the Peyer’s patches and from beneath the intestinal villi in the intestine. In addition there was an associated hypertrophy of the lymph nodes in the ileal loop mesentery. When the intestine was no longer exposed to urine, there was a prompt regeneration and reappearance of the lymphoid elements. The lymphocytes seemed to “home in” on the submucosal areas and reform Peyer’s patches once the urine was removed. The hyperplastic lymph nodes in the ileal loop mesentery regressed when urine exposure was removed. Conventional treatment of ileal loop infections focuses on the upper urinary tracts. Mid-loop stenoses are treated by resection and creation of a new ileal loop conduit. Local therapy in the form of antibiotics and/or steroid enemas may relieve some late midioop obstruction and prevent further damage to renal parenchyma. ACKNOWLEDGMENT We greatly
appreciate
the technical
assistance
provided
by Mr. Paul
Wesley
and
Mr. Steven
Fleit. The animals used in this experiment were superbly cared for by Mr. Henry Syms and other members of the Animal Care Facility of the Children’s Hospital Medical Center. We would like to thank Dr. Arnold Colodny and Dr. Angelo Eraklis for providing specimens from their patients for study. We thank Dr. Fred Rosen for help with interpretation of the histological sections. We appreciate the help of Mrs. Pauline Breen and Mr. Carl Cobb in preparing this manuscript.
REFERENCES I. Seiffert L: Die “Darm-Siphonblase.” fur Klin Chir 183:569. 1935 2. Bricker EM: Bladder
substitution
Arch after
tions, pyelography, renal function, and bacteriology. J Urol Il4:289, 1976 7. Engel RM: Complications of bilateral
pelvic evisceration, Surg Ciin North Am 30: I51 1, 1950 3. Bill AH Jr, Dillard DH, Eggers HE, et al:
uretero-ileal
Urinary and fecal incontinence due to congenital abnormalities in children. Management by implantation of the ureters into an isolated ileostomy. Surg Gynecol Obstet 98:575, 1954 4. Nash DFE: Ileal loop bladder in congeni-
plications of ileal segment urinary diversion in children. Ural Intern 23:97. 1968 9. Logan CW, Scott R Jr, Laskowsky TS: flea1 loop urinary diversion: Evaluation of late results in pediatric urology. J Ural 941544, 1965 10. Ray P, De Demeneco I: Intestinal con-
tal spinal palsy. Br J Ural 28:387, 1956 5. Smith E Durham: Ileo-cutaneous ureterostomy in children. Part 1: Operative technique and complications. Aust NZ J Surg 34:89, 1964 6. Shapiro SR, Lebowitz R, Colodny AH: Fate of 90 children with ileal conduit urinary diversion a decade later: Analysis of complica-
cutaneous
urinary
diversion-a
re-
view of 208 cases. J Ural 101:508, 1969 8. Glenn JF, Small MP, Bovarsky S: Com-
duit urinary diversion in children. Br J Ural 441345, 1972 I I. Retik AB, Perlmutter AD, Gross RE: in children. Cutaneous uretero-ileostomy N Engl J Med 277:217, 1967 12. Smith E Durham: Follow-up studies on
880
TAPPER
150 ileal conduits in children. J Pediatr Surg 7: 1, 1912 13. Colodny, AH: Reconstruction of the urinary system two to fifteen years following establishment of “permanent ileal conduit diversion. J Urol (in press) 14. Wilson WEC, Kirkpatrick CH, Talmage DW: Suppression of immunologic responsiveness in uremia. Ann Intern Med 62: I, 1965 15. Jacobson LO, Simmons EL, Mark EK, et al: Recent studies in recovery from radiation
AND
FOLKMAN
injury after transplantation of heterologous tissue: Effect of Peyer’s patch shielding in irradiated mice. Proc 8th Int Congr Hematol l:8, Sept 1960, Tokyo. Pan Pacific Press, 1962 16. Hardy BE, Lebowitz RL, Baez A, et al: Strictures of the ileal loop. J Ural (in press) 17. Hendren WH: Personal communication 18. Miller-Schoop JW, Good RA: Functional studies of Peyer’s patches: Evidence for their participation in intestinal immune responses. J lmmunol 144: 1757, 1975
Discussion H. Hen&en-(Boston): The long-term results of pediatric ileal loops have been poor in many cases. We have thought that reflux in ileal loops was the main factor in causing recurrent infection and upper tract damage. This work may well implicate other mechanisms. We have not used ileal loops in children for the past 6 yr, turning instead to the use of a nonrefluxing colon conduit that has proven superior. Perhaps the fact that the colon has mucus to protect its mucosa from the effects of urine may be of some importance. We have undiverted 64 children in the past 6 yr, most of them with “permanent” urinary diversion. In 25 of them we had to use the ileal loop to achieve continuity between the kidneys and bladder, because the ureters were short or absent. We have not seen any late strictures in these conduits used as part of a closed urinary tract, which is usually sterile. I wonder, therefore, whether the complication of lymphocyte depletion and stricture is solely the result of urine bathing the mucosa or whether it is the combination of urine plus bacteria. J. Folkman (closing): I want to thank Jay Grosfeld, because 6 yr ago he presented a paper at the academy meetings on the use of the appendix in the dog to replace the ureter. He showed that as time went on this huge lymphoid population in the appendix disappeared when urine poured over it, and then questions came up as to why shouldn’t the ileal loop be very resistant to infection. We have in the dogs had ileal loops put in without urine as controls just to see what happens.
and Judah Folkman
N
1926 Seiffert first described the construction of ileal loop conduits for urinary diversion.’ Bricker in the 1950s popularized the use of the ileal loop as a means of supravesical urinary diversion following exenteration for pelvic malignancy in adults.* Bill and Nash as well as others suggested its usefulness in children with nonmalignant disease of the urinary tract.3l4 Many ileal loop conduits have been constructed in children since 1952. There are numerous early reports and several long-term studies of the results of this surgical procedure in children. 5m” In almost all reports there is a 10Y;P20”,/, incidence of renal deterioration after the construction of the loop. It is not clear why some patients with normal renal function deteriorate after creation of the loop. Although a stoma1 or uretero-ileal stenosis is one obvious cause, in most of these patients such derangement cannot be demonstrated. In addition, only lS’{ of patients with already diminished renal function improve after diversion; the remainder are unchanged or deteriorate further.12 Three patterns of complications stand out: recurrence of urinary tract infection, continued deterioration of renal function, and late stenosis or stricture of the midportion of the ileal loop conduit. We have had extensive experience with the creation of ileal loops at The Children’s Hospital Medical Center. Our policy has been that those children who do not have a neurogenic bladder be considered for urinary dediversion.13 We had the opportunity to study intestinal segments exposed to urine and to compare them to intestinal segments no longer exposed to urine because of diversion. We suggest that the recurrent urinary tract infections, the deteriorating renal function, and the midloop stenosis are related to the same underlying problem, namely a severe depletion of lymphoid elements in the intestine. Our studies indicate that it is urine which causes the lymphoid depletion. We have tested this hypothesis clinically and experimentally. I
METHOD
Clinical Studies. Since 1952. 162 ileal loop diversions have been performed at Children’s Hospital Medical Center. Sixteen of the 162 patients underwent dediversion and reestablishment of urinary continuity, and represent the basis for this study. Histories have been previously reported for I1 of the 16.13 The 16 cases were observed at three stages in their clinical course (Fig. IA).
From the Departments of Surgey. Children’s Hospital Medical Center and The Harvard Medical School, Boston. Mass. and Universily of Cali’omia, San Francisco Medical Center, San Francisco. Calif Presented beJore the 7th Annual Meeting of’ the American Pediatric Surgical Association. Boca Raion. Fla.s April 29-Ma)l I, 1976. Address for reprinr requests: Judah Folkman. M.D.. The Children’s Hospital Medical Center, 300 Longwood Avenue, Boston. Mass. 021 IS. w 1976 bv Gntne & Stratron. Inc. Journal of Pediatric Surgery, Vol. 11, No. 5 (October), 1976
871
TAPPER
872
AND FOLKMAN
STAGE I
STAGE II
STAGE III
STAGE I
STAGE II
STAGE III
A
B Fig. 1. (A) Representation of the three operations performed in children undergoing creation of Real loop. Stage I: formation of standard Real loop; stage II: dediversion and reestablishment of urinary continuity; and stage Ill: excision of defunctionalized intestine. (8) Animal model. Stage I: creation of right uretero-ileal anastomosis with ileal loop; stage II: nephroureterectomy with excision of proximal portion of loop; ond stage Ill: excision of defunctionalized intestine.
Stage I: Creation of the ileal loop. The ureter or pelvic calyx was anastomosed ileum with urine draining through the intestine and exiting through a stoma abdominal wall. Stage II: Reestablishment of urinary continuity. Six reconstructions were
to the terminal on the anterior carried
out
by
ureteroureteral anastomosis. One child had an ileocecal-cystoplasty, one had an ileocystoplasty, one had a renal autotransplantation, one had the loop implanted into the bladder, and one had the loop anastomosed to the distal ureteral stump. Stage III: Excision of dejiinctionalized intestinal segment. At least 1 yr after diversion the defunctionalized ileum was removed, This was done only if it was certain that diversion would not again be necessary, and if the loop was not utilized in the reconstruction.
Experimental
Studies
Eleven mongrel dogs weighing 15 kg were similarly staged (Fig. 1B). A combination of intravenous diabutal and open drop ether was used for anesthesia. segment of terminal ileum approxiStage I: Creation of the ileal loop. An appropriate mately 8 cm from the ileocecal valve was selected for the conduit. The blind end of the loop was closed. The right ureter was anastomosed to the back wall near the blind end of the ileal conduit, using mucosa-to-mucosa interrupted sutures of 5-O chromic catgut. The loop was allowed to lie confortably in front of, or behind the cecum. The open end was brought out through a previously constructed stoma site and matured primarily. Intestinal continuity was reestablished.
LYMPHOID
073
DEPLETION
Two weeks after the first operation the dogs were reanesthetized, the loop catheterized, urine residual measured, and serum creatinine determined. After an interval of 4-12 wk Stage II: Partial excision of the loop and nephroureterectomv. (mean 8 wk) a second laparotomy was performed. At this operation the uretero-ileal anastomosis was inspected. A right nephroureterectomy was performed which included approximately 2576 of the proximal portion of the ileal loop and contained the uretero-ileal anastomosis. The blind end of the loop was closed. Histologic sections were prepared from the entire ileum and stained with hematoxylin and eosin. Stage III: Excision of remainder of ileal loop. After 4-8 wk (mean 5 wk) a third laparotomy was performed. At this operation the remaining segment of the loop was excised. Four animals served as controls. In two dogs intestinal loops were created from a similar portion of the terminal ileum. However, the loop was essentially a defunctionalized portion of intestine without any urine flow. In two animals two loops were created; one loop had a uretero-ileal anastomosis with urine flowing through it: the other loop was defunctionalized intestine.
RESULTS
Clinical Study
Sixteen patients underwent dediversion from their iieal loops and reconstruction of the urinary system. Unfortunately, sections of the intestine were not available at each stage of the procedure in every patient. In addition several patients underwent interval procedures to revise the ileal loop. However, results were identical in all patients studied. Stage I. Figure 2 is a representation of the changes seen in the ileal loop intestinal segments prior to urine exposure, after urine exposure, and after dediversion. Figure 2A depicts the normal Peyer’s patches as seen in the terminal ileum of human intestine. The Peyer’s patches are made up of mature lymphoid elements coalescing to form follicles with healthy germinal centers. There also is a layer of mature lymphocytes beneath the intestinal villi (Fig. 2B). Stage II. The effect of urine exposure on the lymphoid elements in the intestine is evident in Fig. 2C. There is a marked depletion of cellular elements. The follicles are smaller and depleted of mature, active cells. The layer of lymphoid elements usually noted beneath the villi contains only scanty lymphoid cells (Fig. 2D). In the histologic sections there was no evidence of lymphoid dilatation to suggest leakage of lymphocytes. Stage III. The ileal segment remains in situ and defunctionalized for at least 1 yr following dediversion. During this absence of urine, the follicles have regenerated (Fig. 2E). The lymphoid cells regain their plump, healthy appearance. The layer above the lamina propria has been repopulated with lymphoid elements (Fig. 2F). Eosinophils are present in the histologic sections which suggest a reparative process. Experimental Study
Ten of the eleven dogs survived. One dog died of distemper. Cannulation of the loop at 2 wk after the first laparotomy showed prompt urine flow with only a 5-10 cc residual in nine out of ten dogs. One animal had slight stoma1 narrowing but catheterization was easy and the loop contained only 10 cc residual. Dog #ll in which no urine was obtained from the loop will be discussed later.
TAPPER
874
I
STAGE I
STAGE II
STAGE
AND
FOLKMAN
III
ILEUM EXPOSURE
NO URINE
TO
NO URINE
URINE 3
Muscle
:
z c,rvm
Fig. 2. Representation of effect of urine on the lymphoid elements in the ileal loops of children undergoing staged dediversion. (A) Confluence of lymphoid follicles to form peyer’s patches with active germinal centers prior to urine exposure. (B) layer of healthy lymphocytes beneath the intestinal villi prior to urine exposure. (C) Marked depletion of lymphoid elements after exposure to urine. (D) Loss of lymphocytes from beneath the villi with urine exposure. (E) Reappearance of lymphoid follicles with removal of urine. (F) Regeneration of lymphocytes beneath the villi.
Stage I. At the first laparotomy a biopsy was taken from the proximal portion of the loop. This served as control intestine, “before exposure to urine.” The Peyer’s patches in the canine terminal ileum are larger and more numerous than those in the human intestine; they also appear more confluent (Fig. 3A). Most contain large follicles with large, active germinal centers. The lymphocytes making up the Peyer’s patches are plump, mature, and healthy. Just above the lamina propria is a second layer of lymphocytes. This layer is approximately 668 cells thick and is beneath the villi (Fig. 3B). At the villous tips there are more plump, mature lymphocytes. Stage II. After approximately 8 wk, during which time the ileal loop was continuously exposed to urine, the second laparotomy was performed. Histologic sections were made on the entire portion of the resected ileum containing the uretero-ileal anastomosis. The results were similar in all the animals. Figure 3C is representative of the effect of urine on the intestine especially on the lymphoid elements. There is marked depletion of all lymphoid elements espe-
LYMPHOID
STAGE
875
DEPLETION
STAGE III
STAGE II
I
ILEUM NO URINE
Fig. 3. loops. (A)
RE TO
Representation of the effect of urine on the lymphoid elements in the canine ileal Confluent Peyer’s patches as seen in normal canine intestine prior to urine expo-
sure. (B) lymphocytes beneath the intestinal villi prior to urine exposure. (C) Severe lymphoid depletion and toss of follicle architecture with exposure to urine. (D) Depletion of lymphocytes from beneath the villi. (E) Reappearance of lymphoid follicles with active germinal centers with the removal of urine exposure. (F) lymphocytes beginning to reappear beneath intestinal villi as urine exposure is removed.
cially within the follicles. Most of them have been destroyed. The remaining cells are no longer mature and healthy; there are only fine reticular lymphocytes. In the histologic sections there was no evidence of lymphoid dilatation. The lymphoid cell layer just beneath the villi has disappeared and only a few scattered lymphocytes remain (Fig. 3D). Stage ZIZ. After a period of approximately 5 wk, during which time no urine flowed through the loop, the third laparotomy was performed. Figure 3E is a representative section and shows the cellular repopulation taking place. Lymphoid cells are reaccumulating beneath the villi. There are small new follicles forming (Fig. 3F). A moderate number of eosinophils related to the recovery process are also present in the histologic preparations. Figure 4 is from Dog #l which had the longest recovery period (8 wk). Restoration of follicles continues to progress even 8 wk after removal of urine from the loop. More lymphocytes are forming along the border of the crypts and as if by a “homing signal” they are coalescing to form follicles. At the second laparotomy we noticed hypertrophy of the lymph nodes confined to the mesentery of the ileal loop segment (Fig. 5). There was no evidence
876
TAPPER
AND FOLKMAN
Fig. 4. The lymphocytes beneath the villi appear to “home in” and to reform follicles once the urine exposure is removed. This appears to be time dependent.
lymphoid
of a generalized lymphadenopathy. We did not see hypertrophy of mesenteric lymph nodes in the control animals without urine exposure. Hematoxin- eosinstained sections showed reactive hyperplasia of the lymph nodes without any evidence of granuloma formation. When the final portions of ileum were removed at the third laparotomy, the nodes in the mesentery had regressed and were no longer easily visible. The intestine from the control animals showed no evidence of atrophy, ischemia or depletion of lymphoid elements. The number of mature lymphoid elements remained unchanged. Dog #11 which did not make urine through the loop was staged similarly STAGE
I
I
STAGE
II
When urine expoFig. 5. sure is present, there is an associated hypertrophy of the lymph nodes in the mesentery of the ileal loop. The hypertrophic nodes shrink when urine exposure is removed.
LYMPHOID
DEPLETION
a77
Fig. 6. Kidney from dog which did not make x11 urine. Note the large area of segmental infarction in the renal parenchyma.
to the other animals. At the second laparotomy there was no evidence of ureteral obstruction or hydronephrosis. On the contrary the kidney was markedly smaller on the right side. The uretero-ileal anastomosis was patent and only a small amount of methylene blue was noted to come down the ureter. Figure 6 is a section of the involved kidney and shows segmental ischemia. The damage was probably secondary to segmental renal artery stenosis or occlusion which was not detected at the first laparotomy. No dissection was done in the retroperitoneum at the first stage but the possibility of an iatrogenic injury cannot be ruled out. The pathology of the diverted intestinal loop was similar to the control animals without urine exposure. The presence of ureteral epithelium does not alter morphology in the loop. Serum creatinine values were normal. There was no evidence of renal failure in any of the experimental animals. No systemic infections occurred in any of the dogs. Urinalysis at the time of loop catheterizations showed epithelial debris, and a small number of white cells. No colony counts were done. DlSCUSSlON
These studies demonstrate that exposure of intestinal segments to urine causes marked lymphoid depletion in the segments. When the urine exposure is interrupted as by dediversion in patients or nephroureterectomy in dogs, the
878
TAPPER
AND
FOLKMAN
lymphocytes return. These studies give additional support to the idea advanced by others that a lymphocyte suppressive factor exists in urine.r4 If such a factor exists in urine and causes lymphocyte depletion, this may help explain the occurrence of late midloop strictures, recurrent infections, and continued renal deterioration. Peyer’s patches have tremendous regenerative capacity as demonstrated by the return of lymphoid elements after the urine was removed. Jacobsen showed that if 1 Peyer’s patch in a mouse was exteriorized and protected with a lead shield, he could give normally lethal doses of whole body irradiation without fatality.15 The mouse repopulated his entire lymphoid system from 1 protected Peyer’s patch. The continued presence of urine bathing the intestinal mucosa appears to locally inhibit regeneration of the Peyer’s patches. In addition the loss of the lymphocyte barrier in the intestinal segment may predispose the lymph nodes in the ileal loop mesentery to hypertrophy. These hypertrophied nodules may be responsible for late midloop strictures. Hardy has reported 17 cases of late midloop strictures. These occurred in patients several years after construction of their ileal 10op.l~ Loopograms on these patients showed midloop narrowing and stenosis with proximal dilatation. There was no evidence of vascular compromise. All of these loops were removed because of deteriorating renal function. The mesentery of these loops contained hypertrophied lymph nodes which seemed to encroach and narrow the intestinal lumen. This finding of hypertrophied mesenteric nodes has been confirmed by others.” This reactive hyperplasia of the lymph nodes in the heal mesentery may be due to lymphoid depletion in the intestinal segment. The mesenteric nodes would then act as a second line of defense against organisms which invaded the ileal mucosa. The ileal loop segment with its hypertrophied nodes looks similar to intestine involved in regional enteritis. The small intestine is normally resistant to infection. However, the loss of the villous architecture and the depletion of lymphocytes from the Peyer’s patches make the intestinal segment little more than a conduit. Peyer’s patches are immunocompetent lymphoid organs which participate in intestinal immune responses. I8 The depletion of these lymphoid cells coupled with the refluxing nature of the ileal conduit may lead to recurrent infection and continued kidney damage. Current treatment focuses on the upper urinary tracts. Systemic antibiotics are given to decrease bacterial flora in the urine and to protect the kidney from further damage. It may be fruitful for pediatric surgeons and urologists to reexamine conventional treatment for ileal loop infections and for midloop stenosis. Local therapy in the form of nonabsorbable intestinal antibiotics may prevent further damage to renal parenchyma. The conventional treatment for midloop stenosis with proximal-loop dilatation is resection of the loop and construction of a new ileal conduit. However, it may be possible to use a combination of steroid enemas and nonabsorbable antibiotics to shrink hypertrophied mesenteric nodes and relieve obstruction before a permanent stricture develops. This might prevent an additional operation to fashion a new ileal loop.
LYMPHOlD
DEPLETION
a79
SUMMARY
Recent long-term follow-up studies of children undergoing ileal loop urinary diversion have shown a number of complications arising in these patients: recurrent urinary tract infections, deteriorating renal function, and late midloop stricture. We suggest that all three may be related to the same underlying problem, namely, a severe depletion of lymphoid elements in the intestine. In both clinical and animal studies urine exposure to intestinal segments resulted in a marked depletion of lymphoid elements from the Peyer’s patches and from beneath the intestinal villi in the intestine. In addition there was an associated hypertrophy of the lymph nodes in the ileal loop mesentery. When the intestine was no longer exposed to urine, there was a prompt regeneration and reappearance of the lymphoid elements. The lymphocytes seemed to “home in” on the submucosal areas and reform Peyer’s patches once the urine was removed. The hyperplastic lymph nodes in the ileal loop mesentery regressed when urine exposure was removed. Conventional treatment of ileal loop infections focuses on the upper urinary tracts. Mid-loop stenoses are treated by resection and creation of a new ileal loop conduit. Local therapy in the form of antibiotics and/or steroid enemas may relieve some late midioop obstruction and prevent further damage to renal parenchyma. ACKNOWLEDGMENT We greatly
appreciate
the technical
assistance
provided
by Mr. Paul
Wesley
and
Mr. Steven
Fleit. The animals used in this experiment were superbly cared for by Mr. Henry Syms and other members of the Animal Care Facility of the Children’s Hospital Medical Center. We would like to thank Dr. Arnold Colodny and Dr. Angelo Eraklis for providing specimens from their patients for study. We thank Dr. Fred Rosen for help with interpretation of the histological sections. We appreciate the help of Mrs. Pauline Breen and Mr. Carl Cobb in preparing this manuscript.
REFERENCES I. Seiffert L: Die “Darm-Siphonblase.” fur Klin Chir 183:569. 1935 2. Bricker EM: Bladder
substitution
Arch after
tions, pyelography, renal function, and bacteriology. J Urol Il4:289, 1976 7. Engel RM: Complications of bilateral
pelvic evisceration, Surg Ciin North Am 30: I51 1, 1950 3. Bill AH Jr, Dillard DH, Eggers HE, et al:
uretero-ileal
Urinary and fecal incontinence due to congenital abnormalities in children. Management by implantation of the ureters into an isolated ileostomy. Surg Gynecol Obstet 98:575, 1954 4. Nash DFE: Ileal loop bladder in congeni-
plications of ileal segment urinary diversion in children. Ural Intern 23:97. 1968 9. Logan CW, Scott R Jr, Laskowsky TS: flea1 loop urinary diversion: Evaluation of late results in pediatric urology. J Ural 941544, 1965 10. Ray P, De Demeneco I: Intestinal con-
tal spinal palsy. Br J Ural 28:387, 1956 5. Smith E Durham: Ileo-cutaneous ureterostomy in children. Part 1: Operative technique and complications. Aust NZ J Surg 34:89, 1964 6. Shapiro SR, Lebowitz R, Colodny AH: Fate of 90 children with ileal conduit urinary diversion a decade later: Analysis of complica-
cutaneous
urinary
diversion-a
re-
view of 208 cases. J Ural 101:508, 1969 8. Glenn JF, Small MP, Bovarsky S: Com-
duit urinary diversion in children. Br J Ural 441345, 1972 I I. Retik AB, Perlmutter AD, Gross RE: in children. Cutaneous uretero-ileostomy N Engl J Med 277:217, 1967 12. Smith E Durham: Follow-up studies on
880
TAPPER
150 ileal conduits in children. J Pediatr Surg 7: 1, 1912 13. Colodny, AH: Reconstruction of the urinary system two to fifteen years following establishment of “permanent ileal conduit diversion. J Urol (in press) 14. Wilson WEC, Kirkpatrick CH, Talmage DW: Suppression of immunologic responsiveness in uremia. Ann Intern Med 62: I, 1965 15. Jacobson LO, Simmons EL, Mark EK, et al: Recent studies in recovery from radiation
AND
FOLKMAN
injury after transplantation of heterologous tissue: Effect of Peyer’s patch shielding in irradiated mice. Proc 8th Int Congr Hematol l:8, Sept 1960, Tokyo. Pan Pacific Press, 1962 16. Hardy BE, Lebowitz RL, Baez A, et al: Strictures of the ileal loop. J Ural (in press) 17. Hendren WH: Personal communication 18. Miller-Schoop JW, Good RA: Functional studies of Peyer’s patches: Evidence for their participation in intestinal immune responses. J lmmunol 144: 1757, 1975
Discussion H. Hen&en-(Boston): The long-term results of pediatric ileal loops have been poor in many cases. We have thought that reflux in ileal loops was the main factor in causing recurrent infection and upper tract damage. This work may well implicate other mechanisms. We have not used ileal loops in children for the past 6 yr, turning instead to the use of a nonrefluxing colon conduit that has proven superior. Perhaps the fact that the colon has mucus to protect its mucosa from the effects of urine may be of some importance. We have undiverted 64 children in the past 6 yr, most of them with “permanent” urinary diversion. In 25 of them we had to use the ileal loop to achieve continuity between the kidneys and bladder, because the ureters were short or absent. We have not seen any late strictures in these conduits used as part of a closed urinary tract, which is usually sterile. I wonder, therefore, whether the complication of lymphocyte depletion and stricture is solely the result of urine bathing the mucosa or whether it is the combination of urine plus bacteria. J. Folkman (closing): I want to thank Jay Grosfeld, because 6 yr ago he presented a paper at the academy meetings on the use of the appendix in the dog to replace the ureter. He showed that as time went on this huge lymphoid population in the appendix disappeared when urine poured over it, and then questions came up as to why shouldn’t the ileal loop be very resistant to infection. We have in the dogs had ileal loops put in without urine as controls just to see what happens.