British Juurnul of Urolow (1978). 50, 492-495
Radical Cystectomy for Carcinoma of the Bladder P. B. CLARK The General Infirmary and St James’s University Hospital, Leeds
Summary-A personal series of 5 0 radical cystectomies has been reviewed t o decide whether it is a justifiable operation and, if so, when it should be performed. In spite of the fact that radical cystectomy had a higher operative mortality than simple cystectomy and was sometimes followed by lymphoedema, in patients with invaded iliac lymph nodes it was followed by a 25% 5-year survival. It appears, therefore, t o be a justifiable procedure. It is recommended that simple cystectomy should be performed for patients with papillomatosis, carcinoma in situ, and as a salvage procedure after radiotherapy has failed, and that radical cystectomy should be reserved for elective cases of invasive vesical tumours, and for those patients who are found at exploration t o have obvious metastatic deposits in the iliac lymph nodes.
In radical cystectomy the iliac lymph nodes are removed with the bladder. Daughtry et al. (1977) have compared the results of simple and radical cystectomy over a 10-year period at the Cleveland Clinic. They reached the provocative conclusion that radical cystectomy has little justification, firstly because it has a higher operative mortality than simple cystectomy, and secondly because none of their patients with invaded lymph nodes survived 2 years. This study presents the results of 58 cystectomy patients with a possible 5-year survival and it seeks to answer 2 questions: is radical cystectomy justifiable and, if so, in which cases?
Patients and Methods From February 1967 until April 1973, 58 patients have had cystectomy performed for carcinoma of the bladder. Fifty had radical and 8 simple cystectomy. The number of simple cystectomies is too small to analyse in detail. Of the radical cystectomies, 38 were men and 12 women. Eight were aged between 35 and 55, 31 between 55 and 70 and 11 were over 70 years. Half of the patients had elective and half salvage cystectomy, radiotherapy having failed. The indications are shown in Table 1 : elective cystectomy was usually performed for
invasive tumour, occasionally for papillomatosis which was beyond endoscopic control; the commonest indication for salvage cystectomy was persistent bleeding and all of the patients had either residual or recurrent tumours, often multiple and extensive. Pre-operative radiotherapy was given to 7 of the 25 elective radical cystectomy patients. Some were treated with a dose of about 40oO rads in 3 weeks, followed by cystectomy within 4 weeks; others with 2400 rads in 1 week, followed by immediate cystectomy. Cystectomy and urinary diversion were always a one-stage procedure. Five of the earlier cases had a ureterocolic anastomosis and 1 a rectal bladder; the rest had an ileal conduit and this is now the routine method of urinary diversion; the ureters were joined to the ileum end-to-end using the “Y” anastomosis (Clark, 1978). In men, both for simple and radical cystectomy, the seminal Table 1 Indications for Radical Cystectomy Elective
Salvage
Papillomatosis Invasive tumour Ureteric obstruction
14
Total
25
6
Kesidual tumour Recurrent tumour
5 25
(with persistent bleeding
Read at the 34th Annual Meeting of the British Association of Urological Surgeons in Brighton, June 1978.
492
I5 10
16)
493
RADICAL CYSTECTOMY I.OK CARCINOMA 01. THE BLADDER
Table 2 Operative Mortality After Radical Cystectomy Elective
Salvage
Pulmonary oedema Wound infection and sepsis Pulmonary embolus lleus and uraemia Inhaled vomit
1 1
1 1 1 1
Total
2
1
5
vesicles and prostate were removed with the bladder; in women, anterior exenteration was performed. For radical cystectomy, the fat and fascia around the bladder and the lymph node chains from the bifurcation of the common iliac artery to the inguinal ligaments were removed en bloc with the bladder, ligating the arteries at their origin from the internal iliac. The lymph nodes around the common iliac artery were also removed when they were enlarged. The whole length of the urethra was removed in only 3 of the 38 men. Two patients had 1 kidney and its ureter removed en bloc with the bladder and urethra. The operation was almost always performed under epidural or spinal anaesthesia to reduce bleeding, and through a wide transverse lower abdominal incision. A light general anaesthetic was also given to keep the patient asleep through the operation, which usually lasted from 3 to 5 h.
Results There was no operative mortality from simple cystectomy; that from radical cystectomy was 14%. The causes of death are listed in Table 2. Sudden pulmonary oedema killed 2 patients on the second post-operative day and has led to the routine monitoring of central venous pressure. Of the 7 patients who died 5 had had salvage cystectomy, making the operative mortality for Table 3 Late Complications After Radical Cystec-
tomy Ileal conduit Stoma1 stenosis Ureteroileal stenosis Hernia alongside stoma
Ureterocolic anastomosis Hyperchloraemic acidosis
Other 5 1 1
Urethral carcinoma 1 Ureteric carcinoma 3 Intestinal obstruction 2 Ileovaginal fistula 1 Oedema 10
elective radical cystectomy 8% and that for salvage radical cystectomy 20%. Leakage from the ureteroileal anastomosis (1 patient) was treated by bilateral nephrostomy, from the intestinal anastomosis (1 patient) by resuture. One patient developed osteitis pubis. Late complications (Table 3) are divided into those associated with the urinary diversion and other complications. There was a significant incidence of oedema (Table 4). Sometimes the oedema was pitting oedema and transient; sometimes it was nonpitting, permanent and distressing to the patient. Permanent lymphoedema sometimes occurred years after cystectomy and preceded death from carcinomatosis by only a few months; sometimes it persisted for years and, as there was no evidence of recurrent tumour, was presumably caused by damage to lymphatic channels. The cause of death within 5 years in 25 patients was usually carcinomatosis (15 patients); some died from other tumours (2 with ureteric unrelated tumour) and other unrelated causes (2). Two cases have been lost to follow-up, are presumed dead and the cause of death in 3 is unknown. No patient died from renal failure. Five-year survival is shown in Table 5 . After simple cystectomy it was 50%; after radical cystectomy 36%. After elective radical cystectomy it was 28%; after salvage radical cystectomy 44%. These differences can be explained to some extent by the different operative mortality and proportion of Stage T1 or A cases in the 4 groups. When post-operative deaths have been excluded the difference between the 5-year survival after simple and radical cystectomy becomes less-50 and 42% respectively-whereas that between elective and salvage radical cystectomy becomes greater-30 and 55 070 respectively. These corrected 5-year survivals are mirrored to some extent by the percentage of Stage T1 or A cases in the 4 groups. The 5-year survival after radical cystectomy according to the stage of the tumour, both by the TNM and the Jewett classifications, is shown in Table 6. These are uncorrected figures. It is of note that 3 of the 12 patients with lymph node Table 4 Oedema After Radical Cystectomy Pitting 4/43 Transient
(9%) 2
Non-Pitting 6/43 Died within 1 year Survived 6-7 years
(14%) 4
2
494
BRITISH J O U R N A L OF UROLOGY
Table 5 5-Year Survival After Cystectomy Number of cases
Proportion Stage TI / A
14%
50% 42 To
50% 46 070
8 070 20%
30% 55%
24 Vo 72%
Operative mortality
Simple cystectomy Radical cystectomy
50
50% 36%
Elective rad. cystectomy Salvage rad. cystectomy
25 25
44or0
8
Corr. 5-year survival
5-year survival
0QJO
28%
metastases (Stage D,) survived 5 years. Elective radical cystectomy had been performed in all 3 cases and none had received radiotherapy. One had 5 glands involved and another the highest gland in the chain. One remains well 10 years after cystectomy. The cases of lymph node metastasis are analysed further in Table 7. The accuracy of diagnosing lymph node metastasis at the time of operation is assessed in Table 8. Two patients with apparently normal nodes had metastases on histology; both had elective cystectomy for invasive tumours. Fourteen patients with enlarged nodes had no metastases on histology, and the enlargement was presumably an inflammatory response. The 5-year survival for the 7 cases of elective radical cystectomy who received pre-operative radiotherapy was almost identical to that in the 18 cases who received none-29 and 28% respectively.
Discussion There seems little doubt that the operative mortality for radical cystectomy is greater than that for simple cystectomy. Although Whitmore and Marshall (1962) recorded an identical operative mortality of 14% for both simple and radical cystectomy, Cordonnier (1968) recorded one of 4.6% for simple cystectomy. The mortality of the Cleveland Clinic (Daughtry et al., 1977) was 16% for radical and none for simple cystectomyvery similar to that for this series (14 and 0%). Realistic mortality figures would probably be 5 % for simple and 10% for radical cystectomy, and 10% is in fact my own operative mortality for radical cystectomy in the last 5 years.
What are the chances of a patient with lymph node metastasis surviving for 5 years after radical cystectomy? Marshall and Whitmore (1956) decided that if more than 2 lymph nodes were invaded radical surgery was contraindicated; in 1962 they recorded a 4 % 5-year survival. Dretler et al. (1973) excised the whole of the iliac lymph node chains, as in Wertheim’s hysterectomy, and later recorded a 33% 5-year survival for those with only 1 or 2 nodes involved, and a 17% 5-year survival for all cases. Daughtry ef al. (1977) had no long-term survivors; in this series a 25% 5-year survival is recorded for all cases. Ideally, radical cystectomy should be performed only for those cases with invaded lymph nodes. The difficulty is in diagnosis. Lymphography is of limited value with early invasion, and this series has shown the difficulty in deciding at operation which nodes are invaded; even nodes which appear normal may contain tumour. Radical cystectomy has several disadvantages. It has a higher operative mortality than simple cystectomy; it may be followed by lymphoedema; it is more difficult to perform and takes longer. The only reason for performing it, rather than simple cystectomy, is to cure patients with invaded lymph nodes, and the chances of doing this must be weighed against its disadvantages, especially against its higher operative mortality. It is recommended, therefore, that simple cystectomy should be performed routinely for patients with papillomatosis, carcinoma in situ and as a salvage procedure-these patients had a relatively low incidence of lymph node metastasis in this series-whereas radical cystectomy should be reserved for elective cases of invasive tumours of the bladder-which had a high incidence-and for those patients who are found at exploration
Table 6 Radical Cystectorny: 5-Year Survival and Stage TI=A
T2=B,
T3
T4
Unknown
Total
B*
C
D,
495
RADICAL CYSTECTOMY FOR CARCINOMA OF THE BLADDER
Table 7 Radical Cystectomy: Incidence of Lymph Node Metastasis Elective
Salvage
17%
10%
In this series, as in the Cleveland Clinic series, no benefit was found from pre-operative radiotherapy, and the value of this before cystectomy remains sub judice.
Invasive tumours Papillomatosis
References
to have obvious metastatic deposits in the iliac nodes. Salvage cystectomy in this series had a high operative mortality. Many of the patients had been bleeding for months and were poor operative risks. However, the high 5-year survival showed that the procedure was fully justified. Often, when radiotherapy has been given a long time before, it was easier to remove the bladder by dissecting it out near the pelvic wall; it is recommended that lymphadenectomy should be performed only when the nodes are obviously invaded.
Clark, P. B. (1978). End-to-end ureteroileal anastomosis for ileal conduits. British Journal of Urology. (In press.) Cordonnier, J. J. (1968). Cystectomy for carcinoma of the bladder. Journal of Urology. 99, 172-173. Daughtry, J. D., Susan, L. P., Stewart, B. H. and Straffon, R. A. (1977). Ileal conduit and cystectomy; a 10-year retrospective study of ileal conduits performed in conjunction with cystectomy and with a minimum 5-year follow-up. Journal of Urology, 118, 556-557. Dretler, S. P., Ragsdale, B. D. and Leadbetter, W. F. (1973). The value of pelvic lymphadenectomy in the surgical treatment of bladder cancer. Journal of Urology, 109, 414-416. Marshall, V. F. and Whitmore, W. F., Jr. (1956). Surgical treatment of cancers of the urinary bladder. Cancer, 9,
Table 8 Accuracy of Operative Assessment of Lymph Nodes in 41 patients
Whitmore, W. F., Jr. and Marshall, V. F. (1962). Radical total cystectomy for cancer of the bladder; 230 consecutive cases five years later. Journal of Urology, 87, 853-868.
609-619.
Histology of lymph nodes Operativefindings
Normal
Invaded
Nodes normal
16
2
Nodes enlarged
14
9
The Author Philip Clark, MA, M D , MChir, FRCS, Consultant Urological Surgeon, The General Infirmary, Great George Street, Leeds.
Radical Cystectomy for Carcinoma of the Bladder P. B. CLARK The General Infirmary and St James’s University Hospital, Leeds
Summary-A personal series of 5 0 radical cystectomies has been reviewed t o decide whether it is a justifiable operation and, if so, when it should be performed. In spite of the fact that radical cystectomy had a higher operative mortality than simple cystectomy and was sometimes followed by lymphoedema, in patients with invaded iliac lymph nodes it was followed by a 25% 5-year survival. It appears, therefore, t o be a justifiable procedure. It is recommended that simple cystectomy should be performed for patients with papillomatosis, carcinoma in situ, and as a salvage procedure after radiotherapy has failed, and that radical cystectomy should be reserved for elective cases of invasive vesical tumours, and for those patients who are found at exploration t o have obvious metastatic deposits in the iliac lymph nodes.
In radical cystectomy the iliac lymph nodes are removed with the bladder. Daughtry et al. (1977) have compared the results of simple and radical cystectomy over a 10-year period at the Cleveland Clinic. They reached the provocative conclusion that radical cystectomy has little justification, firstly because it has a higher operative mortality than simple cystectomy, and secondly because none of their patients with invaded lymph nodes survived 2 years. This study presents the results of 58 cystectomy patients with a possible 5-year survival and it seeks to answer 2 questions: is radical cystectomy justifiable and, if so, in which cases?
Patients and Methods From February 1967 until April 1973, 58 patients have had cystectomy performed for carcinoma of the bladder. Fifty had radical and 8 simple cystectomy. The number of simple cystectomies is too small to analyse in detail. Of the radical cystectomies, 38 were men and 12 women. Eight were aged between 35 and 55, 31 between 55 and 70 and 11 were over 70 years. Half of the patients had elective and half salvage cystectomy, radiotherapy having failed. The indications are shown in Table 1 : elective cystectomy was usually performed for
invasive tumour, occasionally for papillomatosis which was beyond endoscopic control; the commonest indication for salvage cystectomy was persistent bleeding and all of the patients had either residual or recurrent tumours, often multiple and extensive. Pre-operative radiotherapy was given to 7 of the 25 elective radical cystectomy patients. Some were treated with a dose of about 40oO rads in 3 weeks, followed by cystectomy within 4 weeks; others with 2400 rads in 1 week, followed by immediate cystectomy. Cystectomy and urinary diversion were always a one-stage procedure. Five of the earlier cases had a ureterocolic anastomosis and 1 a rectal bladder; the rest had an ileal conduit and this is now the routine method of urinary diversion; the ureters were joined to the ileum end-to-end using the “Y” anastomosis (Clark, 1978). In men, both for simple and radical cystectomy, the seminal Table 1 Indications for Radical Cystectomy Elective
Salvage
Papillomatosis Invasive tumour Ureteric obstruction
14
Total
25
6
Kesidual tumour Recurrent tumour
5 25
(with persistent bleeding
Read at the 34th Annual Meeting of the British Association of Urological Surgeons in Brighton, June 1978.
492
I5 10
16)
493
RADICAL CYSTECTOMY I.OK CARCINOMA 01. THE BLADDER
Table 2 Operative Mortality After Radical Cystectomy Elective
Salvage
Pulmonary oedema Wound infection and sepsis Pulmonary embolus lleus and uraemia Inhaled vomit
1 1
1 1 1 1
Total
2
1
5
vesicles and prostate were removed with the bladder; in women, anterior exenteration was performed. For radical cystectomy, the fat and fascia around the bladder and the lymph node chains from the bifurcation of the common iliac artery to the inguinal ligaments were removed en bloc with the bladder, ligating the arteries at their origin from the internal iliac. The lymph nodes around the common iliac artery were also removed when they were enlarged. The whole length of the urethra was removed in only 3 of the 38 men. Two patients had 1 kidney and its ureter removed en bloc with the bladder and urethra. The operation was almost always performed under epidural or spinal anaesthesia to reduce bleeding, and through a wide transverse lower abdominal incision. A light general anaesthetic was also given to keep the patient asleep through the operation, which usually lasted from 3 to 5 h.
Results There was no operative mortality from simple cystectomy; that from radical cystectomy was 14%. The causes of death are listed in Table 2. Sudden pulmonary oedema killed 2 patients on the second post-operative day and has led to the routine monitoring of central venous pressure. Of the 7 patients who died 5 had had salvage cystectomy, making the operative mortality for Table 3 Late Complications After Radical Cystec-
tomy Ileal conduit Stoma1 stenosis Ureteroileal stenosis Hernia alongside stoma
Ureterocolic anastomosis Hyperchloraemic acidosis
Other 5 1 1
Urethral carcinoma 1 Ureteric carcinoma 3 Intestinal obstruction 2 Ileovaginal fistula 1 Oedema 10
elective radical cystectomy 8% and that for salvage radical cystectomy 20%. Leakage from the ureteroileal anastomosis (1 patient) was treated by bilateral nephrostomy, from the intestinal anastomosis (1 patient) by resuture. One patient developed osteitis pubis. Late complications (Table 3) are divided into those associated with the urinary diversion and other complications. There was a significant incidence of oedema (Table 4). Sometimes the oedema was pitting oedema and transient; sometimes it was nonpitting, permanent and distressing to the patient. Permanent lymphoedema sometimes occurred years after cystectomy and preceded death from carcinomatosis by only a few months; sometimes it persisted for years and, as there was no evidence of recurrent tumour, was presumably caused by damage to lymphatic channels. The cause of death within 5 years in 25 patients was usually carcinomatosis (15 patients); some died from other tumours (2 with ureteric unrelated tumour) and other unrelated causes (2). Two cases have been lost to follow-up, are presumed dead and the cause of death in 3 is unknown. No patient died from renal failure. Five-year survival is shown in Table 5 . After simple cystectomy it was 50%; after radical cystectomy 36%. After elective radical cystectomy it was 28%; after salvage radical cystectomy 44%. These differences can be explained to some extent by the different operative mortality and proportion of Stage T1 or A cases in the 4 groups. When post-operative deaths have been excluded the difference between the 5-year survival after simple and radical cystectomy becomes less-50 and 42% respectively-whereas that between elective and salvage radical cystectomy becomes greater-30 and 55 070 respectively. These corrected 5-year survivals are mirrored to some extent by the percentage of Stage T1 or A cases in the 4 groups. The 5-year survival after radical cystectomy according to the stage of the tumour, both by the TNM and the Jewett classifications, is shown in Table 6. These are uncorrected figures. It is of note that 3 of the 12 patients with lymph node Table 4 Oedema After Radical Cystectomy Pitting 4/43 Transient
(9%) 2
Non-Pitting 6/43 Died within 1 year Survived 6-7 years
(14%) 4
2
494
BRITISH J O U R N A L OF UROLOGY
Table 5 5-Year Survival After Cystectomy Number of cases
Proportion Stage TI / A
14%
50% 42 To
50% 46 070
8 070 20%
30% 55%
24 Vo 72%
Operative mortality
Simple cystectomy Radical cystectomy
50
50% 36%
Elective rad. cystectomy Salvage rad. cystectomy
25 25
44or0
8
Corr. 5-year survival
5-year survival
0QJO
28%
metastases (Stage D,) survived 5 years. Elective radical cystectomy had been performed in all 3 cases and none had received radiotherapy. One had 5 glands involved and another the highest gland in the chain. One remains well 10 years after cystectomy. The cases of lymph node metastasis are analysed further in Table 7. The accuracy of diagnosing lymph node metastasis at the time of operation is assessed in Table 8. Two patients with apparently normal nodes had metastases on histology; both had elective cystectomy for invasive tumours. Fourteen patients with enlarged nodes had no metastases on histology, and the enlargement was presumably an inflammatory response. The 5-year survival for the 7 cases of elective radical cystectomy who received pre-operative radiotherapy was almost identical to that in the 18 cases who received none-29 and 28% respectively.
Discussion There seems little doubt that the operative mortality for radical cystectomy is greater than that for simple cystectomy. Although Whitmore and Marshall (1962) recorded an identical operative mortality of 14% for both simple and radical cystectomy, Cordonnier (1968) recorded one of 4.6% for simple cystectomy. The mortality of the Cleveland Clinic (Daughtry et al., 1977) was 16% for radical and none for simple cystectomyvery similar to that for this series (14 and 0%). Realistic mortality figures would probably be 5 % for simple and 10% for radical cystectomy, and 10% is in fact my own operative mortality for radical cystectomy in the last 5 years.
What are the chances of a patient with lymph node metastasis surviving for 5 years after radical cystectomy? Marshall and Whitmore (1956) decided that if more than 2 lymph nodes were invaded radical surgery was contraindicated; in 1962 they recorded a 4 % 5-year survival. Dretler et al. (1973) excised the whole of the iliac lymph node chains, as in Wertheim’s hysterectomy, and later recorded a 33% 5-year survival for those with only 1 or 2 nodes involved, and a 17% 5-year survival for all cases. Daughtry ef al. (1977) had no long-term survivors; in this series a 25% 5-year survival is recorded for all cases. Ideally, radical cystectomy should be performed only for those cases with invaded lymph nodes. The difficulty is in diagnosis. Lymphography is of limited value with early invasion, and this series has shown the difficulty in deciding at operation which nodes are invaded; even nodes which appear normal may contain tumour. Radical cystectomy has several disadvantages. It has a higher operative mortality than simple cystectomy; it may be followed by lymphoedema; it is more difficult to perform and takes longer. The only reason for performing it, rather than simple cystectomy, is to cure patients with invaded lymph nodes, and the chances of doing this must be weighed against its disadvantages, especially against its higher operative mortality. It is recommended, therefore, that simple cystectomy should be performed routinely for patients with papillomatosis, carcinoma in situ and as a salvage procedure-these patients had a relatively low incidence of lymph node metastasis in this series-whereas radical cystectomy should be reserved for elective cases of invasive tumours of the bladder-which had a high incidence-and for those patients who are found at exploration
Table 6 Radical Cystectorny: 5-Year Survival and Stage TI=A
T2=B,
T3
T4
Unknown
Total
B*
C
D,
495
RADICAL CYSTECTOMY FOR CARCINOMA OF THE BLADDER
Table 7 Radical Cystectomy: Incidence of Lymph Node Metastasis Elective
Salvage
17%
10%
In this series, as in the Cleveland Clinic series, no benefit was found from pre-operative radiotherapy, and the value of this before cystectomy remains sub judice.
Invasive tumours Papillomatosis
References
to have obvious metastatic deposits in the iliac nodes. Salvage cystectomy in this series had a high operative mortality. Many of the patients had been bleeding for months and were poor operative risks. However, the high 5-year survival showed that the procedure was fully justified. Often, when radiotherapy has been given a long time before, it was easier to remove the bladder by dissecting it out near the pelvic wall; it is recommended that lymphadenectomy should be performed only when the nodes are obviously invaded.
Clark, P. B. (1978). End-to-end ureteroileal anastomosis for ileal conduits. British Journal of Urology. (In press.) Cordonnier, J. J. (1968). Cystectomy for carcinoma of the bladder. Journal of Urology. 99, 172-173. Daughtry, J. D., Susan, L. P., Stewart, B. H. and Straffon, R. A. (1977). Ileal conduit and cystectomy; a 10-year retrospective study of ileal conduits performed in conjunction with cystectomy and with a minimum 5-year follow-up. Journal of Urology, 118, 556-557. Dretler, S. P., Ragsdale, B. D. and Leadbetter, W. F. (1973). The value of pelvic lymphadenectomy in the surgical treatment of bladder cancer. Journal of Urology, 109, 414-416. Marshall, V. F. and Whitmore, W. F., Jr. (1956). Surgical treatment of cancers of the urinary bladder. Cancer, 9,
Table 8 Accuracy of Operative Assessment of Lymph Nodes in 41 patients
Whitmore, W. F., Jr. and Marshall, V. F. (1962). Radical total cystectomy for cancer of the bladder; 230 consecutive cases five years later. Journal of Urology, 87, 853-868.
609-619.
Histology of lymph nodes Operativefindings
Normal
Invaded
Nodes normal
16
2
Nodes enlarged
14
9
The Author Philip Clark, MA, M D , MChir, FRCS, Consultant Urological Surgeon, The General Infirmary, Great George Street, Leeds.
