- Email: [email protected]
Palliative percutaneous and endoscopic urinary diversion for malignant ureteral obstruction
PALLIATIVE PERCUTANEOUS AND ENDOSCOPIC URINARY DIVERSION FOR MALIGNANT URETERAL OBSTRUCTION MARK GASPARINI, M.D. PETER CARROLL, M.D.* MARSHALL STOLLER, M.D. From the Department of Urology, University of California School of Medicine, San Francisco, California
ABS TRA C T--We performed a retrospective analysis of 22 patients with malignant ureteral obstruction who underwent palliative urinary diversion by retrograde ureteral stenting or nephrostomy tube placement. The average duration of survival after diversion was 526 days and was unrelated to tumor type, patient's age or sex, renal ]unction, or indications ]or diversion. As a group, patients without previous hormonal or chemotherapy survived longer. Morbidity related to the urinary diversion was low. The maiority of patients (77 %) were discharged from the hospital, and this group spent 86 percent of their survival time at home. We conclude that modern palliative urinary diversion can be performed with low morbidity and can result in long-term survival and improved quality of life. Predictions or assumptions concerning survival of indi, vidual patients should be made with caution.
Ureteral obstruction secondary to malignancy is a common condition in which the physician must address the feasibility and appropriateness of urinary diversion. A review of the literature reveals little published data. 1 Furthermore, many articles, written several years ago, suggest that urinary diversion in such patients may be associated with high morbidity and mortality rates that may no longer be accurate. 1-6 We provide an update on patients with malignant ureteral obstruction in whom urinary diversion was performed with endourologic techniques within the last three years. Material and Methods From July 1986 to July 1989, 31 patients underwent urinary diversion for malignant ureteral obstruction at our institution. Patients lost to follow-up (5) and those who underwent open *Recipient of an American Cancer Society Clinical Ontology Career Development Award.
408
urinary diversion (4) wel retrospective review. Co formation was available 19 being followed up to t The study group incl women aged thirty to eig 62 years). In all cases the elated with either distant extension significant enou resection. Specifically, lignancies (ovarian, cervi IV, all urologic malignanc were Stage D, and all lignancies (gastric, anal were Stage D. The ma (57 %) had a gynecologic More than half of the pat: for their malignancy be II). The interval betweei sion averaged 79.6 weeks We were able to revie~ 19 patients. At the time ot
UROLOGY
NOVEMBER 1991
/
VOL_ _
•.i--.-i--,t-~.t--r,.~r.~r./3
q.,-,-4,,,,-r,,,~f,~r,,~
~
,,I,---,q---,,.}--r/3r,/3r,/3
r.,/3
,,t.-~
-i---r~r,~
'5 II
II
O
I
I
I
+
+
+
+
+
I
+
+
I
I
+
+
I
+
I
+~
+
I
I=
o I= It
Z k~ o
,..4
g
I -.--.
0
+
+
+
+
I
+
+
+
+
+
+
+'~
+
+
I
+
+
+
,
,
+
+
,
+
,
+
+
+~
+
+
,
,
I "~
+
+
+-~
+
+ e-
AI
g 8 II
o
--~~-~~-~ ~ ~ o o o ~ ~ o
~
~
b" II
l
~ Nlt
~0
$~;:: LOCY
N O V E M B E R 1991
/
V O L U M E XXXVIII, N U M B E R 5
~--
409
'IA~BLE]i Tumor type and history of prior hormonal or chemotherapy q3:Lnlor Type
No.
ii:ervieal i; ast:~:o:intestina! 3tadder (3varian ii~:ostate Lymphoma 3nknown primary TOTALS
5 5 4 3 2 2 1 22
TABLE V.
Prior H or C No. (%) 3 3 1 3 I 1 0 12
(60) (60) (25) (100) (50) (50) (54)
Tumor typ
Tumor Type
No.
Prostate Cervical Gastrointestinal Ovarian Lymphoma Unknown primary tumor Bladder Combined average
2 5 5 3 2 1 4
*Brackets indicate 1 patient still al
KEY: H = hormonal therapy; C = chemotherapy.
TABLE III. Relationship of renal function a~ presentation to survival Serum Creatinine* No. of Survival No. Improved (mg/dL) Patients (Wks) After Drainage <2.0 6 57 (6-168) 5/6 >_2.0 13 88 (1-244) 11/13 >8.0 ~ (of 13) 41 (1-82) 1/2 *Data available in 19 of 22 patients.
TABLEIV. Indications for diversion in 22 patients Indication for E iversion H:ydronephrosis' Renal failuret ~epsis Uremia Flank pain
No. of Patients (%) 22 13 2 2 12
(100) (59) (9) (9) (54)
*Ten patients had bilateral hydronephrosis (46 %). TCreatinine >_2.0 rng/dL.
e:reatinine level was 3.1 mg/dL (normal range, 3.5-1.4 mg/dL). Thirteen patients (59%) had hnpaired renal function (creatinine >2.0 mg/ dL), including 2 with uremia and a creatinine af >8.0 mg/dL. Of the 6 patients with levels _*!.,'2.0 mg/dL, 4 had bilateral hydronephrosis and 1 had unilateral hydronephrosis with urosepsis (Table HI). All patients had documented hydronephrosis on ultrasound (19), computerized tomogral?hy (CT) (1), or excretory urography (2). Ten patients (46 % ) presented with bilateral hydronephrosis and 2 patients (9%) with urosepsis (Table IV). Percutaneous nephrostomy tube placement was performed via a posterolateral puncture with ultrasound guidance and fluoroscopic confirmation in 15 patients (68 % ) and was successful in all. It was the initial procedure in 4 pat:ients (18%), 2 with urosepsis and 2 with uremia, and a secondary procedure in 11 patients (50 %) in whom retrograde ureteral stent placement had been unsuccessful. Ureteral 410
TABLE VI.
Effect of h, chemotherapy on
Previous Therapy
No. Patient:
Yes
13 (59)
No
9 (41)
stents were placed under fll and were successful as an i: cedure in 7 of 18 patients (3! 6F to 8F in diameter; the could be passed was used. 1 a nephrostomy tube was e: ternalized ureteral stent vi proach. In none of our pati be discontinued. A single in appropriate antibiotic was procedures. Results In 22 patients who un urinary diversion for maL struction, survival time aft( eraged seventy-five weeks 0 years). Three patients were the study period. As a g~ prostate cancer survived those with bladder cancer V). The 17 patients who wq the hospital spent, on ave~ their survival time at homl 111 wks); 14 of these pati( provement in their renal available in 3). In the gro percentage of their surviva the hospital was 68 percent Renal function at the tii not correlate with subsequ III). Although not statistic~ absence of previous hormon (N = 9) corresponded wi with a mean survival fir weeks as opposed to fifty-fi
UROLOGY / NOVEMBER 1991 / VOLUI
TABLE VII.
Complications*
,n
No.
No./Pt.
No./Yr. Diverted
22 8
1.6 1.3
1.9 0.5
UTI
~aion
~
....... ts
19 1.5 1.9 22 3.7 1.4 ~ percutaneous nephrostomy. ~!~diaths, cardiac or pulmonary complications, perforation, p or sepsis related to diversion occurred.
~ilhad already received systemic therapy nplications were infrequent and did not g affect survival or quality of life. Febrile iy tract infections (UTI) and stent or irostomy tube occlusion represented the i diversion-related complications (Table None of our patients died as a result of the ~ion procedure, and there were no ear~ pulmonary, septic, or hemorrhagic eom~ti0ns. We observed a total of thirty febrile ~ which represented 1.1 per year of diverLand occurred slightly more frequently in bnts with nephrostomies. Obstruction (or occurred forty-one times and 5 episodes per year of diversion. ral stent changes were scheduled four months, but many patients ~liant. Comment bstruction secondary to wideor abdominal malignancy is not alliative urinary diversion can be pen, endoscopie, or percutaneous ~,ral reports have shown that open 'ies much higher morbidity and ,~s for this group of patients. 2'4-6 studies have shown very high [ mortality rates for diversion perdoseopie or pereutaneous means, many physicians are understandto advise urinary diversion for paadvanced m a l i g n a n t disease. Phee, and Grabstald, z reviewing ~pen nephrostomy diversion in canperformed from 1955 to 1974, reajor life-threatening complication ereent and questioned the value of version in these patients. Keidan et ,~ported a median survival of only I~!~een weeks for 20 patmnts with malignant ~eral obstruction undergoing pereutaneous ~!!rostomy from 1989. to 1986. Forty percent ! heir patients had eareinomatosis, and meL(K;Y / NOVEMBER1991
dian survival for this group was only seven weeks. In our study, we sought to determine the morbidity and mortality of palliative urinary diversion performed by modern endourologie techniques. Our study group represented a limited number of high-risk patients with detailed follow-up. All patients had advanced malignancies, and 86 percent were followed up to their deaths. Open urinary diversions were excluded. All pereutaneous nephrostomy tube placements were performed with ultrasound guidance and fluoroscopic confirmation; all retrograde ureteral stents were placed under fluoroscopic guidance. We found a surprisingly long mean survival time of seventy-five weeks following diversion in our 22 patients. The range was great (1 to 244 weeks) among these patients with similar malignancies and equally advanced disease. As in previous studies, no significant correlation eould be made between length of survival and tumor type, patient's age and sex, renal function, or indications for diversion, s,7,8We noted a tendency for increased length of survival in patients who underwent diversion before receiving treatment for their malignancy (Table VI). This may represent a group of patients whose malignancy could be controlled to some extent by subsequent treatment. We presume that this correlation could become stronger if more effective systemic treatment becomes available. Renal function at the time of presentation did not correlate with post-diversion survival. In fact, 1 patient who presented with uremia and serum ereatinine >8.0 mg/dL went on to live 1.5 years. Renal function after diversion was noteworthy in that all patients with no improvement died within six weeks. Morbidity related to the diversion was low. Unlike previous series, in ours no perioperative deaths occurred nor were there cardiac, pulmonary, or hemorrhagic complications. 2,4-6 Febrile urinary tract infections, the only complication, occurred infrequently (an average of 1/year of diversion) and appeared to be more common with pereutaneous nephrostomy than with ureteral stents. Obstruction or dislodgement occurred only about 1.5 times per year of diversion, suggesting t h a t a routine stent or nephrostomy tube change every four months is adequate. Internal ureteral stenting is often desirable over nephrostomy diversion in terms of patient's comfort and convenience as well as ease
/ VOLUME XXXVIII, NUMBER 5
411
of insertion. T M The low success rate of ureteral stenting in patients with malignant extrinsic ureteral obstruction necessitates nephrostomy drainage in many patients. 12,13 Although difficult to assess objectively, the inconvenience attributed to percutaneous nephrostomy drainage in our group of patients appeared to be minimal. All nephrostomy tubes are placed at the posterior axillary line, and this lateral position affords more comfort for the reclining or supine patient than a more posteriorly placed tube. The majority of patients eared for their nephrostomy independently and were able to bathe or shower without difficulty. All nephrostomies were minimally dressed, and all patients were instructed in the use of leg bags and tube function. A lower initial s u c c e s s rate of ureteral stenting for malignant ureteral obstruction (39 % ) is consistent with previous reports. 12a3 This may be a consequence of marked ureteral tortuosity and kinking of ureters obstructed secondary to extrinsic compression. Our minimal complication rate may be attributed to consistent use of fluoroscopy and the early appreciation of impassable segments with retrograde contrast studies. Once the urologist commits to manipulating an obstructed ureter and finds retrograde drainage unsuccessful, pereutaneous drainage must be seriously considered. Therefore, physicians should be prepared to proceed with percutaneous nephrostomy tube placement in all patients in whom the commitment to diversion has been made. A brief discussion concerning the poor survival and high morbidity of previous reports of palliative urinary diversion is warranted. Many of the previous reports, written many years ago, included a large percentage of patients who underwent open urinary diversion via cutaneous ureterostomies, ileal conduits, open nephrostomies, etc. L5 It is apparent from these studies and others that patients with widespread malignancy do not tolerate or benefit from these open procedures. ~,5,6,9 Similarly, more recent reports on pereutaneous urinary diversion have resulted in progressively lower morbidity and longer survival, possibly as a consequence of refinements in endourologie techniques, advances in chemotherapy, etc. 2,4,6 Keidan e t a l . , 4 who reported the shortest postdiversion survival in the recent literature, studied a group of patients with a 40 percent incidence of eareinomatosis. Such patients may represent a group with very limited survival 412
with or without diversion. None of our had earcinomatosis. The ethical questions surrounding diversion in patients with advanced real must be addressed on an individual basi i ty of life after diversion is difficult to a jeetively. We have shown that patients, vaneed malignancy can undergo p~ urinary diversion with minimal morbi, that the majority of survival time is spel the hospital. In conclusion, we have found that palliatl urinary diversion for patients with m,'di~i~i ureteral obstruction can result in long-term ~ vival with minimal morbidity. One should i~:~i fully consider all patients with malignant} teral obstruction as possible eandidates;i urinary diversion. We believe that a mor6 timistic presentation of the expected outconli palliative urinary diversion can be made t6 tients in light of the results of this study. Hi ever, the great variability in survival a ~ those with similarly advanced disease prec!~]::~ accurate predictions for the individual pa!i{i Department University of San Francisco, California f (DR. S References 1. Grabstald H, and M patient, South Med J 66: i 2. Holden S, McPhee } urinary diversion in eance 3. Dudley S, et al: Per, gynecologic malignancy, ( 4. Keidan R, Greenber: taneous nephrostomy for ] with advanced cancer? Al 5. Sharer W, Grayhael diversion for malignant (1978). 6. Brin EN, Schiff M, a for pelvic malignancy, J [ 7. Zadra J, et ah Nono]~ ureteral obstruction, Can, 8. Soper J, et ah Percut eology patients, Am J Ob~ 9. Andriole G, BettmaI ing Double- I ureteral si urinary drainage, J Urol ] 10. Singh B, Kim H, al there a choice? J Urol 121 11. Hepperlen T, Mard ureteral stent in the cance 12. Doeimo S, and De~ ureteral stents in patients 277 (1989), 13. Khan A, and Utz D prostate associated with bJ
816 (1975).
UROLOGY
/
NOVEMBER 1991
ABS TRA C T--We performed a retrospective analysis of 22 patients with malignant ureteral obstruction who underwent palliative urinary diversion by retrograde ureteral stenting or nephrostomy tube placement. The average duration of survival after diversion was 526 days and was unrelated to tumor type, patient's age or sex, renal ]unction, or indications ]or diversion. As a group, patients without previous hormonal or chemotherapy survived longer. Morbidity related to the urinary diversion was low. The maiority of patients (77 %) were discharged from the hospital, and this group spent 86 percent of their survival time at home. We conclude that modern palliative urinary diversion can be performed with low morbidity and can result in long-term survival and improved quality of life. Predictions or assumptions concerning survival of indi, vidual patients should be made with caution.
Ureteral obstruction secondary to malignancy is a common condition in which the physician must address the feasibility and appropriateness of urinary diversion. A review of the literature reveals little published data. 1 Furthermore, many articles, written several years ago, suggest that urinary diversion in such patients may be associated with high morbidity and mortality rates that may no longer be accurate. 1-6 We provide an update on patients with malignant ureteral obstruction in whom urinary diversion was performed with endourologic techniques within the last three years. Material and Methods From July 1986 to July 1989, 31 patients underwent urinary diversion for malignant ureteral obstruction at our institution. Patients lost to follow-up (5) and those who underwent open *Recipient of an American Cancer Society Clinical Ontology Career Development Award.
408
urinary diversion (4) wel retrospective review. Co formation was available 19 being followed up to t The study group incl women aged thirty to eig 62 years). In all cases the elated with either distant extension significant enou resection. Specifically, lignancies (ovarian, cervi IV, all urologic malignanc were Stage D, and all lignancies (gastric, anal were Stage D. The ma (57 %) had a gynecologic More than half of the pat: for their malignancy be II). The interval betweei sion averaged 79.6 weeks We were able to revie~ 19 patients. At the time ot
UROLOGY
NOVEMBER 1991
/
VOL_ _
•.i--.-i--,t-~.t--r,.~r.~r./3
q.,-,-4,,,,-r,,,~f,~r,,~
~
,,I,---,q---,,.}--r/3r,/3r,/3
r.,/3
,,t.-~
-i---r~r,~
'5 II
II
O
I
I
I
+
+
+
+
+
I
+
+
I
I
+
+
I
+
I
+~
+
I
I=
o I= It
Z k~ o
,..4
g
I -.--.
0
+
+
+
+
I
+
+
+
+
+
+
+'~
+
+
I
+
+
+
,
,
+
+
,
+
,
+
+
+~
+
+
,
,
I "~
+
+
+-~
+
+ e-
AI
g 8 II
o
--~~-~~-~ ~ ~ o o o ~ ~ o
~
~
b" II
l
~ Nlt
~0
$~;:: LOCY
N O V E M B E R 1991
/
V O L U M E XXXVIII, N U M B E R 5
~--
409
'IA~BLE]i Tumor type and history of prior hormonal or chemotherapy q3:Lnlor Type
No.
ii:ervieal i; ast:~:o:intestina! 3tadder (3varian ii~:ostate Lymphoma 3nknown primary TOTALS
5 5 4 3 2 2 1 22
TABLE V.
Prior H or C No. (%) 3 3 1 3 I 1 0 12
(60) (60) (25) (100) (50) (50) (54)
Tumor typ
Tumor Type
No.
Prostate Cervical Gastrointestinal Ovarian Lymphoma Unknown primary tumor Bladder Combined average
2 5 5 3 2 1 4
*Brackets indicate 1 patient still al
KEY: H = hormonal therapy; C = chemotherapy.
TABLE III. Relationship of renal function a~ presentation to survival Serum Creatinine* No. of Survival No. Improved (mg/dL) Patients (Wks) After Drainage <2.0 6 57 (6-168) 5/6 >_2.0 13 88 (1-244) 11/13 >8.0 ~ (of 13) 41 (1-82) 1/2 *Data available in 19 of 22 patients.
TABLEIV. Indications for diversion in 22 patients Indication for E iversion H:ydronephrosis' Renal failuret ~epsis Uremia Flank pain
No. of Patients (%) 22 13 2 2 12
(100) (59) (9) (9) (54)
*Ten patients had bilateral hydronephrosis (46 %). TCreatinine >_2.0 rng/dL.
e:reatinine level was 3.1 mg/dL (normal range, 3.5-1.4 mg/dL). Thirteen patients (59%) had hnpaired renal function (creatinine >2.0 mg/ dL), including 2 with uremia and a creatinine af >8.0 mg/dL. Of the 6 patients with levels _*!.,'2.0 mg/dL, 4 had bilateral hydronephrosis and 1 had unilateral hydronephrosis with urosepsis (Table HI). All patients had documented hydronephrosis on ultrasound (19), computerized tomogral?hy (CT) (1), or excretory urography (2). Ten patients (46 % ) presented with bilateral hydronephrosis and 2 patients (9%) with urosepsis (Table IV). Percutaneous nephrostomy tube placement was performed via a posterolateral puncture with ultrasound guidance and fluoroscopic confirmation in 15 patients (68 % ) and was successful in all. It was the initial procedure in 4 pat:ients (18%), 2 with urosepsis and 2 with uremia, and a secondary procedure in 11 patients (50 %) in whom retrograde ureteral stent placement had been unsuccessful. Ureteral 410
TABLE VI.
Effect of h, chemotherapy on
Previous Therapy
No. Patient:
Yes
13 (59)
No
9 (41)
stents were placed under fll and were successful as an i: cedure in 7 of 18 patients (3! 6F to 8F in diameter; the could be passed was used. 1 a nephrostomy tube was e: ternalized ureteral stent vi proach. In none of our pati be discontinued. A single in appropriate antibiotic was procedures. Results In 22 patients who un urinary diversion for maL struction, survival time aft( eraged seventy-five weeks 0 years). Three patients were the study period. As a g~ prostate cancer survived those with bladder cancer V). The 17 patients who wq the hospital spent, on ave~ their survival time at homl 111 wks); 14 of these pati( provement in their renal available in 3). In the gro percentage of their surviva the hospital was 68 percent Renal function at the tii not correlate with subsequ III). Although not statistic~ absence of previous hormon (N = 9) corresponded wi with a mean survival fir weeks as opposed to fifty-fi
UROLOGY / NOVEMBER 1991 / VOLUI
TABLE VII.
Complications*
,n
No.
No./Pt.
No./Yr. Diverted
22 8
1.6 1.3
1.9 0.5
UTI
~aion
~
....... ts
19 1.5 1.9 22 3.7 1.4 ~ percutaneous nephrostomy. ~!~diaths, cardiac or pulmonary complications, perforation, p or sepsis related to diversion occurred.
~ilhad already received systemic therapy nplications were infrequent and did not g affect survival or quality of life. Febrile iy tract infections (UTI) and stent or irostomy tube occlusion represented the i diversion-related complications (Table None of our patients died as a result of the ~ion procedure, and there were no ear~ pulmonary, septic, or hemorrhagic eom~ti0ns. We observed a total of thirty febrile ~ which represented 1.1 per year of diverLand occurred slightly more frequently in bnts with nephrostomies. Obstruction (or occurred forty-one times and 5 episodes per year of diversion. ral stent changes were scheduled four months, but many patients ~liant. Comment bstruction secondary to wideor abdominal malignancy is not alliative urinary diversion can be pen, endoscopie, or percutaneous ~,ral reports have shown that open 'ies much higher morbidity and ,~s for this group of patients. 2'4-6 studies have shown very high [ mortality rates for diversion perdoseopie or pereutaneous means, many physicians are understandto advise urinary diversion for paadvanced m a l i g n a n t disease. Phee, and Grabstald, z reviewing ~pen nephrostomy diversion in canperformed from 1955 to 1974, reajor life-threatening complication ereent and questioned the value of version in these patients. Keidan et ,~ported a median survival of only I~!~een weeks for 20 patmnts with malignant ~eral obstruction undergoing pereutaneous ~!!rostomy from 1989. to 1986. Forty percent ! heir patients had eareinomatosis, and meL(K;Y / NOVEMBER1991
dian survival for this group was only seven weeks. In our study, we sought to determine the morbidity and mortality of palliative urinary diversion performed by modern endourologie techniques. Our study group represented a limited number of high-risk patients with detailed follow-up. All patients had advanced malignancies, and 86 percent were followed up to their deaths. Open urinary diversions were excluded. All pereutaneous nephrostomy tube placements were performed with ultrasound guidance and fluoroscopic confirmation; all retrograde ureteral stents were placed under fluoroscopic guidance. We found a surprisingly long mean survival time of seventy-five weeks following diversion in our 22 patients. The range was great (1 to 244 weeks) among these patients with similar malignancies and equally advanced disease. As in previous studies, no significant correlation eould be made between length of survival and tumor type, patient's age and sex, renal function, or indications for diversion, s,7,8We noted a tendency for increased length of survival in patients who underwent diversion before receiving treatment for their malignancy (Table VI). This may represent a group of patients whose malignancy could be controlled to some extent by subsequent treatment. We presume that this correlation could become stronger if more effective systemic treatment becomes available. Renal function at the time of presentation did not correlate with post-diversion survival. In fact, 1 patient who presented with uremia and serum ereatinine >8.0 mg/dL went on to live 1.5 years. Renal function after diversion was noteworthy in that all patients with no improvement died within six weeks. Morbidity related to the diversion was low. Unlike previous series, in ours no perioperative deaths occurred nor were there cardiac, pulmonary, or hemorrhagic complications. 2,4-6 Febrile urinary tract infections, the only complication, occurred infrequently (an average of 1/year of diversion) and appeared to be more common with pereutaneous nephrostomy than with ureteral stents. Obstruction or dislodgement occurred only about 1.5 times per year of diversion, suggesting t h a t a routine stent or nephrostomy tube change every four months is adequate. Internal ureteral stenting is often desirable over nephrostomy diversion in terms of patient's comfort and convenience as well as ease
/ VOLUME XXXVIII, NUMBER 5
411
of insertion. T M The low success rate of ureteral stenting in patients with malignant extrinsic ureteral obstruction necessitates nephrostomy drainage in many patients. 12,13 Although difficult to assess objectively, the inconvenience attributed to percutaneous nephrostomy drainage in our group of patients appeared to be minimal. All nephrostomy tubes are placed at the posterior axillary line, and this lateral position affords more comfort for the reclining or supine patient than a more posteriorly placed tube. The majority of patients eared for their nephrostomy independently and were able to bathe or shower without difficulty. All nephrostomies were minimally dressed, and all patients were instructed in the use of leg bags and tube function. A lower initial s u c c e s s rate of ureteral stenting for malignant ureteral obstruction (39 % ) is consistent with previous reports. 12a3 This may be a consequence of marked ureteral tortuosity and kinking of ureters obstructed secondary to extrinsic compression. Our minimal complication rate may be attributed to consistent use of fluoroscopy and the early appreciation of impassable segments with retrograde contrast studies. Once the urologist commits to manipulating an obstructed ureter and finds retrograde drainage unsuccessful, pereutaneous drainage must be seriously considered. Therefore, physicians should be prepared to proceed with percutaneous nephrostomy tube placement in all patients in whom the commitment to diversion has been made. A brief discussion concerning the poor survival and high morbidity of previous reports of palliative urinary diversion is warranted. Many of the previous reports, written many years ago, included a large percentage of patients who underwent open urinary diversion via cutaneous ureterostomies, ileal conduits, open nephrostomies, etc. L5 It is apparent from these studies and others that patients with widespread malignancy do not tolerate or benefit from these open procedures. ~,5,6,9 Similarly, more recent reports on pereutaneous urinary diversion have resulted in progressively lower morbidity and longer survival, possibly as a consequence of refinements in endourologie techniques, advances in chemotherapy, etc. 2,4,6 Keidan e t a l . , 4 who reported the shortest postdiversion survival in the recent literature, studied a group of patients with a 40 percent incidence of eareinomatosis. Such patients may represent a group with very limited survival 412
with or without diversion. None of our had earcinomatosis. The ethical questions surrounding diversion in patients with advanced real must be addressed on an individual basi i ty of life after diversion is difficult to a jeetively. We have shown that patients, vaneed malignancy can undergo p~ urinary diversion with minimal morbi, that the majority of survival time is spel the hospital. In conclusion, we have found that palliatl urinary diversion for patients with m,'di~i~i ureteral obstruction can result in long-term ~ vival with minimal morbidity. One should i~:~i fully consider all patients with malignant} teral obstruction as possible eandidates;i urinary diversion. We believe that a mor6 timistic presentation of the expected outconli palliative urinary diversion can be made t6 tients in light of the results of this study. Hi ever, the great variability in survival a ~ those with similarly advanced disease prec!~]::~ accurate predictions for the individual pa!i{i Department University of San Francisco, California f (DR. S References 1. Grabstald H, and M patient, South Med J 66: i 2. Holden S, McPhee } urinary diversion in eance 3. Dudley S, et al: Per, gynecologic malignancy, ( 4. Keidan R, Greenber: taneous nephrostomy for ] with advanced cancer? Al 5. Sharer W, Grayhael diversion for malignant (1978). 6. Brin EN, Schiff M, a for pelvic malignancy, J [ 7. Zadra J, et ah Nono]~ ureteral obstruction, Can, 8. Soper J, et ah Percut eology patients, Am J Ob~ 9. Andriole G, BettmaI ing Double- I ureteral si urinary drainage, J Urol ] 10. Singh B, Kim H, al there a choice? J Urol 121 11. Hepperlen T, Mard ureteral stent in the cance 12. Doeimo S, and De~ ureteral stents in patients 277 (1989), 13. Khan A, and Utz D prostate associated with bJ
816 (1975).
UROLOGY
/
NOVEMBER 1991