Volume 70 March 1977
183
Section of Urology President W Barr Stirling FRCS
with Section of Obstetrics & Gynaecology President R B K Rickford FRCOG
Meeting 22 April 1976
The Pelvic Ureter Mr W F Hendry (St Bartholomew's Hospital, London ECIA 7BE)
This complication is most likely to occur in the presence of dense adhesions - as may be produced by pelvic inflammatory disease or endometriosis (Fig 2) - or when the normal anatomy of the pelvis
Injuries to the Pelvic Ureter
The pelvic ureter is usually injured either at a high level, near the pelvic brim, where it lies adjacent to the ovarian vessels, or low down beside the cervix, where it is crossed by the uterine vessels (Fig 1). The ureter may be included in ligatures, or crushed in clamps and subsequently undergo necrosis; or it may simply be partially or completely divided. bladder
ureter
uterine artery
uterus
ovarian vessels ureter
Fig 1 The uretei-s are most vulnerable w here they lie vessels by
to the ovarian vessels, and low down where they, adjacent are crossed the uterine
Fig 2 Endometriosis involving the left ureter
184
Proc. roy. Soc. Med. Volume 70 March 1977
is distorted, for example by fibroids or ovarian tumours growing into the broad ligament, which may displace the ureter from its normal position. Occasionally the ureter may become devitalized and slough after extensive pelvic dissection, as in Wertheim's hysterectomy, due to interference with its blood supply, particularly after previous pelvic irradiation. Very rarely, the extreme lower end of the ureter may be caught in the high stitches inserted beside the cervix in pelvic floor repair or vaginal hysterectomy. In practice, two situations are likely to arise: the gynecologist may suspect or recognize that the ureter has been injured at the time of operation; or evidence of such damage may become manifest later, either as an urgent or non-urgent complication. In either circumstance there should be no hesitation in seeking expert urological assistance. The complication may be suspected or recognized immediately or recognized later; in the latter event it may be urgent or non-urgent. If the ureter is injured during operation, or if there are grounds for suspecting that it may have been crushed, divided or included in a ligature, the entire length of the pelvic ureter is exposed, and the nature and extent of the injury is defined. This may frequently be facilitated by stripping the peritoneum from the side wall of the pelvis on that side, identifying the ureter, and following it down to the bladder. If injury to the extreme lower end of the ureter is suspected, the bladder may be opened by an anterior cystotomy, and the appropriate ureteric orifice is catheterized. If the catheter
B
i.
iAmk6.r--.
.l..
.1 W-I I,:
ureteric splint
Fig 4 A, injury to left ureter (involved with endomeiriosis) recognized immediately and repaired; B, intravenous urogram 2 months later
Fig 3 High ureteric injuries are best repaired by end-toend anastomosis over a splint. An extraperitoneal drain is advisable
ascends easily and no injury is found, the bladder is simply closed with two layers of chromic catgut, and a small Malecot suprapubic tube or urethral catheter is left in for 10-14 days. The bladder usually heals well and nothing is lost. If, on the other hand, ureteric injury is found, reconstruction or reimplantation can proceed without delay, using tissues that are fresh and easily identified. The choice of procedure for high injuries is end-toend anastomosis, and for low injuries ureteric reimplantation with or without a psoas hitch or Boari flap. If a high ureteric injury is found, direct end-toend anastomosis is performed (Fig 3). The circum-
Section of Urology with Section of Obstetrics and Gynacology
ference of the cut ends of the ureters is increased by a 1-cm vertical incision, and they are then joined using interrupted fine chromic catgut sutures (not silk). This anastomosis is best done over a ureteric splint. An infant's esophageal feeding tube (Portex No. 8 FG) passed up from the bladder is ideal for this purpose, and it has a Luer connexion at its distal end for joining up to the bedside apparatus. Some surgeons prefer a T-tube, but this necessarily involves a second hole in the ureter. A corrugated drain is laid down to the anastomosis extraperitoneally, and removed after four or five days. The ureteric splint is usually removed after about ten days (Figs 4A, B). If a low ureteric injury is found, it is usually possible to reimplant the ureter into the bladder. The lower end of the ureter is dissected out, and may be reimplanted directly into the nearest part of the bladder. Alternatively, the bladder may be opened, and the ureter reimplanted into the posterior or lateral wall. Whichever technique is used, care should be taken to produce a submucosal course of at least 2 cm to prevent reflux. The most important point is to ensure that the ureteric reimplantation is done without tension. This is best achieved by suturing the apex of the bladder to the psoas muscle on that side (the psoas hitch) (Fig 5). The anastomosis may be splinted as previously described, and the bladder is drained by suprapubic or urethral catheter for 2-3 weeks. After radiotherapy or in the presence of dense adhesions, it may not be possible to elevate the peritoneum sufficiently well to perform the ureteric reconstruction extraperitoneally. In these circumstances, a transperitoneal Boari flap procedure may be preferable (Fig 6): the bladder flap is rolled into a tube and joined to the lower end of the ureter
Fig 5 Reimplantation of ureter with reflux preventing submucosal tunnel; note psoas hitch to prevent tension on the anastomosis
185
Fig 6 Boariflap operation; this technique is particularly useful after pelvic irradiation
over a splint. This procedure has the advantage of using well vascularized bladder wall to bridge the gap to the lower end of the ureter, and this is particularly important after previous pelvic irradiation.
Fig 7 42-year-old patient with vaginal leakage of urine 3 months after hysterectomy. Intravenous urogram reported as showing right ureterovaginalfistula: in fact, she had a
vesicovaginalfistula
186
Proc. roy. Soc. Med. Volume 70 March 1977
Ureteric injuries may present after operation ina an urgent or non-urgent manner. If the damagedfl ureter continues to drain urine into the extraperitoneal tissues, pelvic cellulitis may follow; if it drains into the peritoneal cavity, urinary peri-* tonitis will be produced; or if both ureters are injured anuria may result. On the other hand, ligation of one ureter may produce little more than transient loin pain. Probably the most common delayed presentation is by the development of a fl urinary fistula, presenting usually per vaginam, or occasionally through the wound. When evidence of ureteric damage becomes evi- F Ascending ureterogram showing typical dent after operation, the precise nature of the aFig r9earances oJflow ureterovaginalfistula injury should be defined before deciding on the pp best approach to repair. In urgent cases, investigations may be performed as an emergency; in other cases, careful evaluation of the lesion may proceed more slowly. The minimum investigations are urine culture, blood urea and electrolytes, and a high-dose intravenous urogram is essential; it will nearly always indicate the nature of the complication, although great care must be taken to confirm its exact location (Fig 7). This is best done by cystoscopy and ascending ureterogram (Figs 8 and 9). Once the nature and exact site of the lesion have been defined, a decision on the best form of management can be reached. In some early cases, simple bladder or ureteric drainage with a whistle- Fig 10 This low left ureterovaginalfistula healed tip or flute-ended catheter may allow a small fistula spo)ntaneously with catheter drainage
.
.
Fig 8 Ascending ureterogram showing urinary extravasation at a high (ovarian) level
.
,'..,'
.4 . .
-'
.W.
Fig 11 Intravenous urogram 3 months ajter reimplantation of left ureter for ureterovaginalfistula. note eJject ofjpsoas-hitch procedure
Section of Urology with Section of Obstetrics and Gynaccology to heal (Fig 10). More often, however, some form of operative intervention is necessary. If a ureter is completely obstructed, reconstruction or reimplantation should be undertaken within three weeks if worth-while kidney function is to be preserved. On the other hand, if a vaginal fistula develops there is less urgency about the repair, and it is probably worth trying simple drainage for a few days. However, if the fistula shows no sign of drying up, ureteric repair may be undertaken without further delay. There does not seem to be the same necessity for waiting with ureteric injuries as there is with vesicovaginal fistula, where a threemonth delay is required to ensure that the tissues are firm enough to hold the repairing sutures. The operative approach to the repair of ureteric injuries proceeds along similar lines to those outlined above. The ureter is usually best approached through an oblique extraperitoneal incision, and either repaired or reimplanted, with or without a psoas hitch or Boari flap. Urinary drainage should be checked by postoperative pyelogram (Fig 1 1). If a considerable defect has to be bridged, it may occasionally be necessary to do a high transuretero-ureterostomy, or even interpose an isolated loop of ileum between ureter and bladder. Seldom if ever should the need arise for tying off a ureter, nephrectomy, or ureterocolic anastomosis.
Acknowledgment: I am grateful to Miss Freda Wadsworth for the artwork for Figs 1, 3, 5 and 6.
Professor John P Blandy and Mr John D Anderson (The London Hospital, London El ]BB)
Management of the Injured Ureter In a review of 30 cases of ureteric injury (Table 1) the largest group were, predictably, those following pelvic operations, especially when there was unusual difficulty from previous radiotherapy, extensive tumour, or endometriosis. Mr Hendry (p 183) has described the methods by which today an accurate and precise diagnosis may be made by early high-dose urography and bulb ureterography, and we would endorse his plea that the urologist should be sent for as soon as possible in order that these investigations may be done without delay. Once the diagnosis has been made, there is today a wide range of different surgical options (Table 2), so that it should hardly ever be necessary to divert the urine into the sigmoid, and the sacrifice of a good kidney should only be contemplated in those patients who are too extensively affected by pelvic tumour, too frail, or too ill, to warrant a recon-
187
Table 1 Etiology of injuries to the pelvic ureter
Hysterectomy (radiotherapy) Hysterectomy (benign) Abdominoperineal excision Small-bowel obstruction Iliofemoral bypass Operations for ureteric stone Fractures of pelvis Pyeloplasty Vesicoplasty
No. of
No. of
cases
ureters
5 9 5 1 1 4 2 1 2
6 10 7 I I 4 2 1 3
30
35
Table 2 Methods used for primary ureteric repair in 30 patients
Nephrectomy (non-function 3, vascular graft 1) Attempted ureteroneocystostomy Reimplant with psoas hitch Boari-Ockerblad flap
No. of
No. of
cases
ureters
4 4 1
4
4-
21 (+ 2)
I 26 (+ 2)
30 (+2)
35 (+2)
* All failed: 2 later underwent successful Boari procedures, 2 had nephrectomy
structive operation. In addition, the presence of an arterial prosthesis more or less completely contraindicates a nearby urinary anastomosis. No two cases are quite alike, and one must choose how best to save the kidney and restore continuity in the circumstances of each individual case. Ligature of the ureter should never be done: it often leaks soon afterwards and it may well give rise to bacteremia and pyonephrosis on the obstructed side. It may be tempting to try a direct reimplantation of the ureter into the bladder, and this is certainly made more easy by the assistance of a psoas hitch which may remove any tension from the anastomosis; but in most of our cases, what seemed to be a short gap at the outset soon became a lengthy one after the ureter had been liberated from its surroundings and adequately mobilized. It then becomes very difficult to effect an adequate anti-reflux reimplantation without at the same time running the risk of having tension on the anastomosis. Probably because of such tension all four cases in this series came to grief, perhaps because when they were done we were unaware of the advantages of the psoas hitch. The most generally servicable solution to the problem posed by the low injury to the ureter is the Boari-Ockerblad flap (Fig 1) (Boari 1894, Ockerblad 1947) which we have now used in 35 ureters in 30 patients as a primary operation, and in 2 of those in whom an attempted ureteroneocysto-
183
Section of Urology President W Barr Stirling FRCS
with Section of Obstetrics & Gynaecology President R B K Rickford FRCOG
Meeting 22 April 1976
The Pelvic Ureter Mr W F Hendry (St Bartholomew's Hospital, London ECIA 7BE)
This complication is most likely to occur in the presence of dense adhesions - as may be produced by pelvic inflammatory disease or endometriosis (Fig 2) - or when the normal anatomy of the pelvis
Injuries to the Pelvic Ureter
The pelvic ureter is usually injured either at a high level, near the pelvic brim, where it lies adjacent to the ovarian vessels, or low down beside the cervix, where it is crossed by the uterine vessels (Fig 1). The ureter may be included in ligatures, or crushed in clamps and subsequently undergo necrosis; or it may simply be partially or completely divided. bladder
ureter
uterine artery
uterus
ovarian vessels ureter
Fig 1 The uretei-s are most vulnerable w here they lie vessels by
to the ovarian vessels, and low down where they, adjacent are crossed the uterine
Fig 2 Endometriosis involving the left ureter
184
Proc. roy. Soc. Med. Volume 70 March 1977
is distorted, for example by fibroids or ovarian tumours growing into the broad ligament, which may displace the ureter from its normal position. Occasionally the ureter may become devitalized and slough after extensive pelvic dissection, as in Wertheim's hysterectomy, due to interference with its blood supply, particularly after previous pelvic irradiation. Very rarely, the extreme lower end of the ureter may be caught in the high stitches inserted beside the cervix in pelvic floor repair or vaginal hysterectomy. In practice, two situations are likely to arise: the gynecologist may suspect or recognize that the ureter has been injured at the time of operation; or evidence of such damage may become manifest later, either as an urgent or non-urgent complication. In either circumstance there should be no hesitation in seeking expert urological assistance. The complication may be suspected or recognized immediately or recognized later; in the latter event it may be urgent or non-urgent. If the ureter is injured during operation, or if there are grounds for suspecting that it may have been crushed, divided or included in a ligature, the entire length of the pelvic ureter is exposed, and the nature and extent of the injury is defined. This may frequently be facilitated by stripping the peritoneum from the side wall of the pelvis on that side, identifying the ureter, and following it down to the bladder. If injury to the extreme lower end of the ureter is suspected, the bladder may be opened by an anterior cystotomy, and the appropriate ureteric orifice is catheterized. If the catheter
B
i.
iAmk6.r--.
.l..
.1 W-I I,:
ureteric splint
Fig 4 A, injury to left ureter (involved with endomeiriosis) recognized immediately and repaired; B, intravenous urogram 2 months later
Fig 3 High ureteric injuries are best repaired by end-toend anastomosis over a splint. An extraperitoneal drain is advisable
ascends easily and no injury is found, the bladder is simply closed with two layers of chromic catgut, and a small Malecot suprapubic tube or urethral catheter is left in for 10-14 days. The bladder usually heals well and nothing is lost. If, on the other hand, ureteric injury is found, reconstruction or reimplantation can proceed without delay, using tissues that are fresh and easily identified. The choice of procedure for high injuries is end-toend anastomosis, and for low injuries ureteric reimplantation with or without a psoas hitch or Boari flap. If a high ureteric injury is found, direct end-toend anastomosis is performed (Fig 3). The circum-
Section of Urology with Section of Obstetrics and Gynacology
ference of the cut ends of the ureters is increased by a 1-cm vertical incision, and they are then joined using interrupted fine chromic catgut sutures (not silk). This anastomosis is best done over a ureteric splint. An infant's esophageal feeding tube (Portex No. 8 FG) passed up from the bladder is ideal for this purpose, and it has a Luer connexion at its distal end for joining up to the bedside apparatus. Some surgeons prefer a T-tube, but this necessarily involves a second hole in the ureter. A corrugated drain is laid down to the anastomosis extraperitoneally, and removed after four or five days. The ureteric splint is usually removed after about ten days (Figs 4A, B). If a low ureteric injury is found, it is usually possible to reimplant the ureter into the bladder. The lower end of the ureter is dissected out, and may be reimplanted directly into the nearest part of the bladder. Alternatively, the bladder may be opened, and the ureter reimplanted into the posterior or lateral wall. Whichever technique is used, care should be taken to produce a submucosal course of at least 2 cm to prevent reflux. The most important point is to ensure that the ureteric reimplantation is done without tension. This is best achieved by suturing the apex of the bladder to the psoas muscle on that side (the psoas hitch) (Fig 5). The anastomosis may be splinted as previously described, and the bladder is drained by suprapubic or urethral catheter for 2-3 weeks. After radiotherapy or in the presence of dense adhesions, it may not be possible to elevate the peritoneum sufficiently well to perform the ureteric reconstruction extraperitoneally. In these circumstances, a transperitoneal Boari flap procedure may be preferable (Fig 6): the bladder flap is rolled into a tube and joined to the lower end of the ureter
Fig 5 Reimplantation of ureter with reflux preventing submucosal tunnel; note psoas hitch to prevent tension on the anastomosis
185
Fig 6 Boariflap operation; this technique is particularly useful after pelvic irradiation
over a splint. This procedure has the advantage of using well vascularized bladder wall to bridge the gap to the lower end of the ureter, and this is particularly important after previous pelvic irradiation.
Fig 7 42-year-old patient with vaginal leakage of urine 3 months after hysterectomy. Intravenous urogram reported as showing right ureterovaginalfistula: in fact, she had a
vesicovaginalfistula
186
Proc. roy. Soc. Med. Volume 70 March 1977
Ureteric injuries may present after operation ina an urgent or non-urgent manner. If the damagedfl ureter continues to drain urine into the extraperitoneal tissues, pelvic cellulitis may follow; if it drains into the peritoneal cavity, urinary peri-* tonitis will be produced; or if both ureters are injured anuria may result. On the other hand, ligation of one ureter may produce little more than transient loin pain. Probably the most common delayed presentation is by the development of a fl urinary fistula, presenting usually per vaginam, or occasionally through the wound. When evidence of ureteric damage becomes evi- F Ascending ureterogram showing typical dent after operation, the precise nature of the aFig r9earances oJflow ureterovaginalfistula injury should be defined before deciding on the pp best approach to repair. In urgent cases, investigations may be performed as an emergency; in other cases, careful evaluation of the lesion may proceed more slowly. The minimum investigations are urine culture, blood urea and electrolytes, and a high-dose intravenous urogram is essential; it will nearly always indicate the nature of the complication, although great care must be taken to confirm its exact location (Fig 7). This is best done by cystoscopy and ascending ureterogram (Figs 8 and 9). Once the nature and exact site of the lesion have been defined, a decision on the best form of management can be reached. In some early cases, simple bladder or ureteric drainage with a whistle- Fig 10 This low left ureterovaginalfistula healed tip or flute-ended catheter may allow a small fistula spo)ntaneously with catheter drainage
.
.
Fig 8 Ascending ureterogram showing urinary extravasation at a high (ovarian) level
.
,'..,'
.4 . .
-'
.W.
Fig 11 Intravenous urogram 3 months ajter reimplantation of left ureter for ureterovaginalfistula. note eJject ofjpsoas-hitch procedure
Section of Urology with Section of Obstetrics and Gynaccology to heal (Fig 10). More often, however, some form of operative intervention is necessary. If a ureter is completely obstructed, reconstruction or reimplantation should be undertaken within three weeks if worth-while kidney function is to be preserved. On the other hand, if a vaginal fistula develops there is less urgency about the repair, and it is probably worth trying simple drainage for a few days. However, if the fistula shows no sign of drying up, ureteric repair may be undertaken without further delay. There does not seem to be the same necessity for waiting with ureteric injuries as there is with vesicovaginal fistula, where a threemonth delay is required to ensure that the tissues are firm enough to hold the repairing sutures. The operative approach to the repair of ureteric injuries proceeds along similar lines to those outlined above. The ureter is usually best approached through an oblique extraperitoneal incision, and either repaired or reimplanted, with or without a psoas hitch or Boari flap. Urinary drainage should be checked by postoperative pyelogram (Fig 1 1). If a considerable defect has to be bridged, it may occasionally be necessary to do a high transuretero-ureterostomy, or even interpose an isolated loop of ileum between ureter and bladder. Seldom if ever should the need arise for tying off a ureter, nephrectomy, or ureterocolic anastomosis.
Acknowledgment: I am grateful to Miss Freda Wadsworth for the artwork for Figs 1, 3, 5 and 6.
Professor John P Blandy and Mr John D Anderson (The London Hospital, London El ]BB)
Management of the Injured Ureter In a review of 30 cases of ureteric injury (Table 1) the largest group were, predictably, those following pelvic operations, especially when there was unusual difficulty from previous radiotherapy, extensive tumour, or endometriosis. Mr Hendry (p 183) has described the methods by which today an accurate and precise diagnosis may be made by early high-dose urography and bulb ureterography, and we would endorse his plea that the urologist should be sent for as soon as possible in order that these investigations may be done without delay. Once the diagnosis has been made, there is today a wide range of different surgical options (Table 2), so that it should hardly ever be necessary to divert the urine into the sigmoid, and the sacrifice of a good kidney should only be contemplated in those patients who are too extensively affected by pelvic tumour, too frail, or too ill, to warrant a recon-
187
Table 1 Etiology of injuries to the pelvic ureter
Hysterectomy (radiotherapy) Hysterectomy (benign) Abdominoperineal excision Small-bowel obstruction Iliofemoral bypass Operations for ureteric stone Fractures of pelvis Pyeloplasty Vesicoplasty
No. of
No. of
cases
ureters
5 9 5 1 1 4 2 1 2
6 10 7 I I 4 2 1 3
30
35
Table 2 Methods used for primary ureteric repair in 30 patients
Nephrectomy (non-function 3, vascular graft 1) Attempted ureteroneocystostomy Reimplant with psoas hitch Boari-Ockerblad flap
No. of
No. of
cases
ureters
4 4 1
4
4-
21 (+ 2)
I 26 (+ 2)
30 (+2)
35 (+2)
* All failed: 2 later underwent successful Boari procedures, 2 had nephrectomy
structive operation. In addition, the presence of an arterial prosthesis more or less completely contraindicates a nearby urinary anastomosis. No two cases are quite alike, and one must choose how best to save the kidney and restore continuity in the circumstances of each individual case. Ligature of the ureter should never be done: it often leaks soon afterwards and it may well give rise to bacteremia and pyonephrosis on the obstructed side. It may be tempting to try a direct reimplantation of the ureter into the bladder, and this is certainly made more easy by the assistance of a psoas hitch which may remove any tension from the anastomosis; but in most of our cases, what seemed to be a short gap at the outset soon became a lengthy one after the ureter had been liberated from its surroundings and adequately mobilized. It then becomes very difficult to effect an adequate anti-reflux reimplantation without at the same time running the risk of having tension on the anastomosis. Probably because of such tension all four cases in this series came to grief, perhaps because when they were done we were unaware of the advantages of the psoas hitch. The most generally servicable solution to the problem posed by the low injury to the ureter is the Boari-Ockerblad flap (Fig 1) (Boari 1894, Ockerblad 1947) which we have now used in 35 ureters in 30 patients as a primary operation, and in 2 of those in whom an attempted ureteroneocysto-
