Caliceal Fistula Formation Following Renal Transplantation: Management With Partial Nephrectomy and Ureteral Replacement

0022-5347/95/1533-0612$03.00/0 THEJOURVAL OF UROLOGY Copyright 0 1995 by AMERICAN UROMK;ICAL ASSOCIATION,INC

Vol. 153,612-614,March 1995 Printed in U.S.A.

CALICEAL FISTULA FORMATION FOLLOWING RENAL TRANSPLANTATION: MANAGEMENT WITH PARTIAL NEPHRECTOMY AND URETERAL REPLACEMENT J. L. GUTIERREZ-CALZADA, J. RAMOS-TITOS, J. A. GONZALEZ-BONILLA, A. SANTOS GARCIA-VAQUERO, A. MARTIN-MORALES AND R. BURGOS-RODRIGUEZ From the Department

of

Urology and Transplant Unit, Hospital Regional de Malaga, Malaga, Spain

ABSTRACT

Six patients in this series of 543 renal transplants (1.10%) suffered a post-transplant renal segmental infarct of the donor kidney because of occlusion of a n accessory renal artery. Five grafted kidneys had multiple renal arteries. Patients presented with symptoms of a caliceal fistula and were treated by partial (25 to 40%) transplant nephrectomy, followed by closure and tissue coverage with either parietal peritoneum (4patients) or lyophilized human dura mater sealed with fibrin (2). In 2 cases the renal ischemia and necrosis involved the ureter, and a pyelo-pyelostomy was performed. One patient died of cardiorespiratory complications immediately postoperatively. Five years postoperatively all kidneys functioned well without recurrence of fistula and 5 patients returned to a normal life-style. The combination of radical excision and tissue closure, plus ureteral substitution when needed was a n effective treatment that prevented loss of the graft. KEYWORDS:kidney transplantation,kidney calices, fistula, nephrectomy, ureter

Segmentary renal infarction follows ischemia produced by been corrected so that the kidney was acceptable for transan obstruction of a independent polar or segmentary branch plantation. We were later proved wrong because the previof the renal artery. If the necrotic process of an extensive ously poorly perfused pole became necrotic and a urinary infarct penetrates to a calix it causes extravasation of urine fistula developed postoperatively. and leads to a caliceal-cutaneous fistula. Although this cirIn all of these cases extravasated urine began to flow from cumstance is rare (approximately 3%)according to the liter- the drainage tube between 9 and 17 days postoperatively. ature,' when it occurs in a transplanted kidney in which the The daily volumes increased within several days to overtake ischemia is extensive and deep, the patient may lose finally that of the bladder catheter. In some patients, urine the grafted kidney or even die.233 The seriousness of the seeped from the skin wound, and they had low grade fever problem is proportional to the area of kidney surface in- and a slight increase in serum creatinine level. Bacterial farcted and to the amount of urine extravasated. Small cultures of urine from the fistula and voided from the bladder infarcted zones with small urinary flows sometimes seal were positive. We treated the infection according to the themselves by "dry cure" after insertion of an indwelling bacterial sensitivity test but we chose the least nephrodouble pigtail reno-vesical ~ a t h e t e r . ~ - ~ toxic agent. Four patients supported this level of infection We report the treatment of 6 cases in which 25 to 40%of well but 2 had higher levels that approached the threshold the organ was visibly necrotic and associated with the ex- of sepsis and the infections only subsided after surgical travasation of large quantities of urine. In 2 of these cases the removal of the focus of infection. ischemia had extended to cause necrosis of the ureter and Because the caliceal fistulas occurred soon after transplantonly an open operation offered a successful outcome. ation there was great risk of rejection. Therefore, we only decreased immunosuppressive therapy in 1 of the 2 patients MATERIALS AND METHODS whose poor general condition prompted us to lessen the dose We review 6 cases of renal caliceal fistula (1.10% of a of corticosteroids for 2 days and restore it to normal after series of 543 renal transplant patients) caused by a wide- surgical correction of the fistula. The presence of extravaspread segmentary infarct of 25 to 40% of the renal mass. sated urine made the diagnosis of a urinary fistula obvious In 2 cases the extensive ischemia had caused necrosis of but it was always confirmed by biochemical analysis. Echogthe ureter but not the renal pelvis. Because of the great raphy, Doppler ultrasound and computerized tomography scarcity of donor kidneys we attempted to ensure survival were helpful but surgical exploration always revealed the of the intact remainder of these extensively infarcted true extent of the renal infarct and allowed urine leakage grafts when we noted that the kidney was producing suf- points to be identified. It was also easy to determine whether ficient urine to ensure that the patient would not again the ureter was involved. When the presence of a large urinary fistula was confirmed become dependent on dialysis. These segmentary infarcts are the consequence of an oc- partial transplant nephrectomy was done. The infarcted recluded aberrant renal polar or accessory artery in the hilum nal tissue excised in these cases represented 25 to 40%, of the of grafted kidneys with multiple renal arteries that had been kidney. The open calices were closed individually by polyglysubjected to either bench surgery repairs or individual anas- colic acid plus lactic acid braided sutures and then, in the 3 tomoses of the donor arteries with the vessels of the recipi- cases in which the parietal peritoneum was apparently unent.'-'" One donor kidney in this series had only 1 renal affected by extravasated urine, the cut surface was covered artery and despite the fact that 1 pole was poorly perfused, by a pedicled flap of parietal peritoneum,". l 2 which was then we re-perfused it ex situ and the situation appeared to have sutured to the margins of the cut surface. In the other 3 cases the cut parenchymal surface was covered by applying closely Accepted for publication August 19, 1994 and fixing with sutures a layer of lyophilized human dura 612

CALICEAL FISTULA FORMATION AFTER RENAL TRANSPLANTATION

613

eratively). There was no recurrent urinary fistula, and the patients continued to lead a normal life. They could not be differentiated from any other renal transplant patient who had no postoperative surgical complications (see table). Postoperatively, 2 patients had extravasation of small amounts of urine via the drainage tube but this diminished to zero within 3 weeks. In the other 3 cases, the closed urinary fistula gave no further problem (see table). Regular followup examinations of these patients for 5 years showed that the evolution pattern could not be differentiated from that of any other renal transplant patient. Levels of serum creatinine were normal or near normal except when some transient rejection crisis occurred. The mean levels of serum creatinine at 5 years for each patient are shown in the table.

A

DISCUSSION

A, partial nephrectomy of transplanted kidney with substitutionof necrotic ureter by ureter of nephrectomized ipsilateral native kidney. B , substitution of necrotic ureter by native contralateral ureter because of severity of inflammatory process of surrounding tissues irritated by extravasated urine.

mater, which was then sealed with fibrin tissue sealant. In 2 patients urine was drained internally to the bladder for temporary protection using a double pigtail catheter. In the other 4 cases urine was diverted externally via a nephrostomy catheter. All catheters were kept in place for 5 weeks and were not withdrawn until radiological examination indicated that the fistula was completely ~ l o s e d . * , ~ , ~ , ' ~ In the 2 most extensively infarcted transplanted kidneys (40%) the necrotic process also affected the ureter and we had to substitute the ureter of the transplanted kidney with native ureter. We used the native contralateral ureter in 1 case in which extravasated urine had contaminated the ipsilateral native ureter. In each case pyelo-pyelostomy was done between the proximal part of the pelvis of the transplanted kidney and t h e distal portion of the pelvis of the recipient kidney (see figure). We placed a Jackson-Pratt perirenal drainage tube postoperatively in all 6 patients, with no need for suction. RESULTS

Of the original 6 patients 5 enjoyed good renal function (1 died of cardiopulmonary complications immediately postop-

In 5 of the 6 patients the complications were the sequelae of an obstruction of a donor kidney aberrant artery with multiple arteries that were not all included in the Carrel aortic patch. In these situations we generally choose to convert the several arteries into a single arterial trunk by bench surgery. We anastomose terminolaterally 1 or more aberrant arteries to the principal arterial trunk. However, as in 1 patient in this small series, when we find a polar vessel that is too short and too far from the hilum to reach the principal renal artery but that is of sufficient caliber we anastomose it independently to a receptor vessel. The results of our revascularizations using this procedure in transplant kidneys with multiple renal arteries that were not all included in the aortic patch have been generally good. The complications reported occurred hours or days postoperatively despite the fact that we had confirmed, after unclamping the vessels, that renal circulation of the entire organ was normal and that the aberrant vessel pulsed well. Thrombosis of the transplanted aberrant vessel produced an infarct and necrosis of the renal tissues that it supplied. The parenchyma of a segment of a necrotic kidney with a caliceal fistula must be excised completely. Although we close with a suture each open calix exposed on the bleeding surface of the healthy parenchyma, the surface must be covered and reinforced by other tissue. Our procedure is based on experience in trauma surgery of fragmented kidneys of normal (nontransplanted) persons, during which we used with success either a pedicled vascularized flap of parietal peritoneum or lyophilized human dura mater, depending on the anatomical situation in each case. In our 6 cases we first attempted tissue coverage with a flap of vascularized parietal

Review of cases Approximate Extent of Renal Necrotic Process (%)

Situation in

FGB

? 30

Superior pole

EGM

2 30

Superior pole

ACV

r40

Inferior 50%

JUR

2 25

Superior pole

AAG

2 40

Inferior pole

FRO

225

Inferior pole

Pt.

Graft

Immediate postop, Operation

Result

Partial transplant nephrectomy, peritoneum plastic operation, temporary pigtail nephrostomy Partial transplant nephrectomy, dura mater plastic operation, sealed with fibrin, temporary pigtail nephrostomy Partial transplant nephrectomy, dura mater plastic operation, ureter substitution, temporary D o u b l e 2 catheter Partial transplant nephrectomy, dura mater plastic operation, sealed with fibrin, temporary pigtail nephrostomy Partial transplant nepbrectomy, dura mater plastic operation, ureter substitution (native contralat.),temporary vesicorenal Double4 catheter Partial transplant nephrectomy, peritoneum plastic operation, temporary pigtail nephrostomy ~~

* Medical Engineering Corp., New York, New York.

Result /5-Yr. Followup Levels (mg.ldl.1of Serum Creatinine (range)

No extravasation of urine

Complete closure of fistula/ 1.4(0.9-2.2)

No extravasation of urine

Complete closure of fistula/ 0.9 (0.9-1.7)

Urinary leakage for 3 wks.

Complete closure of fistula/ 1.7(1.2-2.5)

Patient died immediately poStoP. No extravasation of urine

Not applicable

Urinary leakage for 2 wks.

Complete closure of fistula/ 2.3 (l.W.1)

Complete closure of fistula/ 2.5 (1.7-2.9)

614

CALICEAL FISTULA FORMATION AFTER RENAL TRANSPLANTATION

peritoneum that we apply closely to the cut surface of the healthy parenchyma and suture around the periphery. However, in some sectioned transplanted kidneys the peritoneal tissues that surround the transplant were swollen and irritated by extravasated urine, and could not be used for tissue coverage. In these cases we covered the parenchymal surface with a free implant of a layer of human dura mater and sealed the periphery with fibrin. It might appear that the ideal solution for an extensively necrotic kidney would be to remove it, return the patient to dialysis and await another transplant to avoid the life threatening situation.2.'2 However, we were faced with the facts that the number of kidney donors was seriously insufficient and that some of these patients could not be given another compatible transplant because a hyperimmune state had developed. In addition, we found that, despite the fact that large areas of parenchyma were necrotic the affected kidneys in these cases were functioning well, with acceptable levels of serum creatinine that we reasoned would improve once the extravasation of urine and the superimposed infection had been treated. CONCLUSION

This combination of partial nephrectomy followed by tissue coverage of the sectioned kidney with either parietal peritoneum or with lyophilized human dura mater sealed with fibrin, and substitution of the transplant ureter by pyelopyelostomy between the proximal pelvis of the transplanted kidney and the distal pelvis of a recipient kidney had saved those organs that might have been considered a total loss. Our experience with these 6 cases demonstrated that transplanted kidneys tolerated further intrinsic surgery just as well as native kidneys. Our results suggest that it will always be worth the effort to save a functioning partially necrotic transplanted kidney. Mr. D. W. Schofield provided helpful suggestions while editing and translating the manuscript.

REFERENCES

1. Tilney, N. L. and Kirkman, R. L.: Surgical aspects of kidney transplantation. In: Renal Transplantation. Edited by M. R. Garovoy and R. D. Guttmann. New York: Churchill Livingstone, chapt. 5, p. 93, 1986. 2. Mundy, A. R., Podesta, M. L., Bewick, M., Rudge, C. J. and Ellis, F. G.: The urological complications of 1000 renal transplants. Brit. J. Urol., 5 3 397, 1978. 3. Loughlin, K R., Tilney, N. L. and Richie, J. P.: Urologic complications in 718 renal transplant patients. Surgery, 95 297, 1984. 4. Schiff, M., Jr., McGuire, E. J., Weiss, R. M. and Lytton, B.: Management of urinary fistulas after renal transplantation. J. Urol., 115 251, 1976. 5. Berger, R. E., Ansell, J. S., Tremann, J. A., Herz, J. H., Fbttazzi, L. C. and Marchioro, T. L.: The use of self-retained ureteral stents in the management of urologic complications in renal transplant recipients. J. Urol., 124 781, 1980. 6. Streem, S. B., Novick, A. C., Steinmuller, D. R. and Nusselman, P. W.: Percutaneous techniques for the management of urological renal transplant complications. J. Urol., 135: 456, 1986. 7. Schiff, M., Jr., McGuire, E. J., Weiss, R. M. and Lytton, B.: Management of urinary fistulas after renal transplantation. J. Urol., 115 251, 1976. 8. Salvatierra, O., Jr., Olcott, C., IV,Amend, W. J.,Jr., Cochrum, K C. and Feduska, N. J.: Urological complicationsof renal transplantation can be prevented or controlled. J. Urol., 117 421, 1977. 9. Novick, A. C., Magnusson, M. and Braun, W. E.: Multiple-artery renal transplantation: emphasis on extracorporeal methods of donor arterial reconstruction. J. Urol., 122: 731, 1979. 10. GutiBrrez-Calzada, J. L., Ramos-Tito, J., GonzAlez-Bonilla,J. A, Burgos-Rodriguez, R., Gonzalez-Molina Alcaide, M., Sancez-Bernal, C. and Allona-Moncada, A.: Transplante renal: anomalias vasculares y procedimientos de revascularizaci6n. Acta Urol. Esp., 5: 265, 1982. 11. Turner-Warwick, R.: The use of the omental pedicle graft in urinary tract reconstruction. J. Urol., 116 341, 1976. 12. Goldstein, L., Cho, S. I. and Olsson, C. A.: Nephrostomy drainage for renal transplant complications. J. Urol., 126 159, 1981,