0022-534 7/92/14 74-1020$03.00/0 THE JOURNAL OF UROLOGY Copyright© 1992 by AMERICAN UROLOGICAL ASSOCIATION, INC.
Vol. 147, 1020-1023, April 1992
Printed in U.S.A.
BACILLUS CALMETTE-GUERIN THERAPY FOR SUPERFICIAL BLADDER CANCER: A 10-YEAR FOLLOWUP HARRY W. HERR, DAVID D. WARTINGER, WILLIAM. R. FAIR AND HERBERT F. OETTGEN From the Urologic Service, Department of Surgery and Clinical Immunology Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, New York
ABSTRACT
We report the outcome of 61 patients with superficial bladder tumors who received bacillus --Galme-tte-Gu-e-r-in (B8G}the-rapy--and -were followed for at lea-st 10-ye-ars (range 10-to 13 years)-;--A
total of 19 patients (31 %) remains free of tumor and progression, 17 (28%) had interval superficial recurrences but no progression and 25 (41 %) had progression, first identified as muscle invasion in 12, prostatic involvement in 8 or metastasis in 5. Most tumors recurred or progressed within the first 5 years. Of the 61 patients 33 (54 %) are disease-free and with an intact bladder, 13 (21 %) are alive after cystectomy, 2 (3%) died of other causes, 1 (2%) is alive with metastasis and 12 (20%) died of metastatic urothelial cancer (11 from the bladder or prostate and 1 from an upper tract tumor). KEY WORDS:
bladder neoplasms; BCG vaccine; carcinoma, transitional cell
lntravesical bacillus Calmette-Guerin (BCG) has been shown in randomized clinical trials to alter the pattern of recurrence1· 2 and progression 3 of superficial bladder tumors. However, such studies do not provide absolute followup of more than 5 years and, therefore, they do not address the question of whether BCG simply delays or actually prevents disease progression. In other words, can any intravesical agent abrogate tumor progression and preserve the bladder, or will ultimate sacrifice of the bladder (and perhaps lives) be an inexorable consequence of the neoplastic diathesis? We report the outcome of high risk bladder cancer patients treated by transurethral resection and BCG for intervals of more than 10 years. MATERIAL AND METHODS
A total of 86 patients with recurrent superficial bladder cancer from 1978 to 1981 was entered into a previously reported prospective controlled clinical trial. 2 All patients had papillary noninvasive (stage Ta) or superficially invasive (stage Tl) tumors (grades 2 and 3), with or without associated carcinoma in situ (Tis). The initial objective was to determine the effect of BCG treatment on the recurrence of superficial bladder cancer after transurethral resection. Of the patients 43 were randomized to receive transurethral resection plus BCG and 43 underwent transurethral resection alone. Of the control (transurethral resection) patients 18 were crossed over to the BCG arm after 6 months to 1 year because of frequent tumor recurrences. These 61 ECG-treated patients have been followed for a minimum: of 10 years and are the focus of this report. BCG (Armand Frappier substrain) was administered intravesically (120 mg. in 50 ml. saline) and percutaneously (5 X 10 7 units). Treatment was begun 2 to 4 weeks after complete transurethral resection of overt papillary lesions and continued weekly for 6 weeks. BCG treatments were repeated intravesically at the same induction dosage and schedule. None of the patients received maintenance BCG. Patients were followed every 3 to 4 months thereafter with cystoscopy, biopsy or transurethral resection and urine cytology. After 5 years disease-free patients were generally evaluated at 6-month intervals. End points selected for analysis included number of patients with and without superficial (stages Ta, Tis and Tl) recurrences, number of patients with progression and the interval to Accepted for publication August 16, 1991. Supported by National Institutes of Health Grants CA-19267 and CA-74146.
progression, and survival from the initial BCG treatment. Tumor progression was defined as muscle invasion (stage T2+ ), metastasis (M+) or prostatic involvement (stage T4p). Prostatic disease included noninvasive urethral and/or ductal carcinoma and stromal invasion. RESULTS
Patient characteristics are listed in table 1. Of the 61 patients 34 (56%) had stage Ta and 27 (44%) had stage Tl tumors. A total of 44 patients (72%) had coexisting carcinoma in situ (Tis) proved by biopsy and positive urine cytology, and they received BCG therapeutically. The other 17 patients underwent resection of papillary lesions only, and they had negative posttransurethral resection cytology studies and received prophylactic BCG. Of the patients 41 received 1 course of BCG, and 20 received 2 (14 patients) or 3 (6 patients) 6-week treatments. The average interval between BCG treatments was 26 months. Sequential BCG courses (that is 12 consecutive weeks) were not given. A total of 19 patients (31 %) remained free of recurrence and progression for more than 10 years after transurethral resection and 1 course of BCG: 1 had stage Ta, 6 had stage Ta + Tis, 1 had stage Tl and 11 had stage Tl + Tis disease. Another 17 TABLE
1. Patient characteristics
No. pts. Median age (range) No. men/No. women Initial T stage (No. pts.): Ta Ta+ Tis Tl Tl+ Tis No. BCG treatments (No. pts.): 1 2 3
61 63 (34-77) 47/14 11 23 6
21 41 14 6
No. pts. without tumor progression (%):
No recurrences Recurrences Type of progression (% ): T2+
T4 (p) M+ Total No.(%) Followup (yrs.): Minimum Median (range)
1020
19 17
(31) (28)
12 8
(20) (13)
5
(8)
25
(41)
10 11.9 (10-13)
1021
BACILLUS CALMETTE-GUEIUN IN SUPERFICIAL BLADDER CANCER
patients (28%) had recunent superficial bladder tumors but no progression during the 10-year period. These patients received BCG initially for stage Ta tumors in 6, stage Ta + Tis in 5 and stage Tl + Tis in 6. The outcome of these 36 patients without progression is shown in table 2. The number of patients with positive cystoscopic evaluations showing recurrent tumor indicates that tumor recurred most often within the initial 5 years (the majority were seen within 2 years). Eight patients had recurrence(s) before but none after 5 years, 2 had no recurrences within 5 years only to suffer recurrent disease at 6 and 7 years, and 7 had sporadic recurrences during the entire 10+-year period. The trend was toward fewer recurrences among fewer patients after 5 and up to 10 years of followup. Recurrences were generally easily controlled by transurethral resection, which was supplemented in some patients with carcinoma in situ by additional BCG. Cystectomy was not done in any patient for recurrent superficial bladder tumors alone. Only 1 patient in this group died of metastatic renal pelvis cancer. There was no difference in the initial T stage among patients who did or did not have superficial tumor recurrence. Progression occurred in 25 patients (41 %): 12 (20%) with muscle invasion, 8 (13%) with prostatic involvement and 5 (18%) with distant metastasis (table 1). The outcome of patients with such progression is shown in table 3. The distribution in initial T stages (stage Ta in 5, stage Ta + Tis in 10, stage Tl in 4 and stage Tl + Tis in 6) did not differ from those in patients without progression. Tumor progressed most often within 5 years, although 3 patients had muscle infiltrating tumors at 6, 8 and 9 years, respectively. No tumor progressed after 10 years. Of the 25 patients 8 (32%) had superficial bladder recurrences at the time or within 6 months before disease progression was documented. The other 13 patients (52%) had de novo progression as specified that was not heralded by a pattern of prior recurrence. Of the 12 patients with stage T2+ tumors 4 had superficial lesions 3 to 6 months previously, and in 3 this was a stage Tl tumor. Radical cystectomy (transurethral resection in 1 patient) salvaged all but 1 patient with progression only in the bladder or prostate. None of the patients with metastasis (with or without local disease) survived beyond 5 years despite chemotherapy. A muscle infiltrating (stage T2+) tumor developed in 12 patients at an average of 33 months after BCG therapy. Four patients had metastases at the time or within 3 months of detection of the bladder cancer and 1 had metastases from an invasive bladder tumor after 8 years. The average interval to stage T2+ progression in this group was 26 months. Seven patients had no distant disease when the stage T2+ bladder tumor occurred. The average interval to invasion among these patients was 45 months. All patients remain disease-free after salvage therapy (table 3). Table 4 shows the outcome of 8 patients with progression in the prostate: prostatic transitional carcinoma was diagnosed within 5 years in 5, and 3 had disease at 5, 7 and 9 years, respectively. The presence or absence of concomitant bladder tumor recurrence was a poor predictor of subsequent prostatic disease, since the bladder was free of disease in 6 patients when the prostatic tumor was documented. One patient with stromal TABLE 2.
Superficial Bladder Re-
No. Pts.
No. Pts. With Recurrences Between
invasion died but the other 7 (5 with noninvasive ductal tumor and 2 with stromal invasion) were salvaged with cystoprostatectomy. Table 5 shows characteristics of the 12 patients who died of urothelial cancer. All 12 patients died of metastases emanating from the bladder (11) or upper tract (1). Five patients had metastases without an invasive bladder tumor, although 4 had had prior superficial recurrences. The other 7 patients had concomitant local and systemic progression, which in 5 was not heralded by a pattern of recurrent superficial tumors. If we assume that distant metastases appearing within 2 years of BCG therapy were from the original bladder tumor (4 patients), then 7 of these patients might have been saved if cystectomy had been performed initially. DISCUSSION
Assessing the impact of conservative management on tumor recurrence, progression and survival in patients with superficial bladder tumors is compromised by the often long natural history of the disease, the competing mortality in the age groups concerned and the lack of studies providing long-term followup. Can some patients be cured, or at least maintain a functioning bladder for more than 10 years? Are enough patients with progression salvaged to justify the initial conservative approach? Of the 61 patients followed for at least 10 and up to 13 years 33 (54%) have a functioning bladder (albeit half required local therapy for tumor recurrence), 13 (21 %) are disease-free after cystectomy, 2 (3%) are dead of unrelated causes, 1 (2%) is currently alive with metastatic disease and 12 (20%) died of metastatic bladder/prostate (11) or upper tract (1) urothelial cancer. Progression of bladder cancer was noted more often within the initial 5 years, suggesting that surveillance should be particularly intense during this interval even among patients without superficial recurrence. Tumor progression after 5 years and up to 10 years was sufficiently common to make lifelong surveillance for panurothelial cancer a necessity in such patients. A total of 11 patients (18%) had upper tract tumors from 1.5 to 9 years after BCG treatment: in 2 metastasis originated from a renal pelvis tumor, while 9 were salvaged with either distal ureterectomy or nephroureterectomy. These patients will be the subject of a separate report. The initial T stage and the presence or number of subsequent superficial bladder tumor recurrences did not correlate with progression. An average of 12 cystoscopy studies identified an average of 38 tumors among patients with superficial recurrences but no progression compared to 5 positive cystoscopy studies showing 14 superficial tumors among patients with progression. Furthermore, because so many of the patients (72%) had associated in situ carcinoma the impact of multifocal disease upon recurrence and progression rates cannot be well evaluated. Why did these high risk patients have a progression rate of less than 50% during 10 years, a high rate of salvage when nonmetastatic progression occurred and a favorable overall survival? An explanation might be selection of patients with tumors having a long natural history and a low rate of metas-
Outcome of patients without progression Av. No. Pos. Cystoscopy Studies (Av. No. Tumors) Between
currences
0-5 Yrs.
Neg. Pos. Totals(%)
19 17
36 (59)
15
5-10 Yrs. 9
0-5 Yrs. 5.7 (20)
* Cerebrovascular accident at 9 years and myocardial infarction at 6 years.
t Metastasis from upper tract tumor.
5-10 Yrs. 3.2 (7)
No. Pts. Disease-Free at More Than 10 Yrs.
No. Pts. Dead of Dis-
No. Pts. Dead of Other Causes
No. Pts. Alive With Disease
ease
16 16
0
32 (52)
I
1t
2* 0
2
1t 0
I
1022
HERR AND ASSOCIATES TABLE 3.
No. Pts. ( %)
Type of Progression
T2+ (muscle invasion) T4p (prostatic involvement) M+ (metastasis) Totals(%)
Outcome of patients with progression
No. Pts. With Progression Between
Av. Mos. to Progression
0-5 Yrs.
5-10 Yrs.
No. Pts. With Bladder Recurrences 0-6 Mos. Before Progression
12 (20)
33
10
2
4
8 (13)
47 20
5 5
3
2 2
20
5
5 (8)
25 (41)
0
No. Pts. With No Evidence of Disease
Treatment of Progression Radical cystectomy Transurethral resection Systemic chemotherapy Radical cystectomy Systemic chemotherapy
8 (32)
6 1 0 7 0 14(56)
Av. Overall Survival (yrs.)*
No. Pts. Dead of Disease
11
10.8 3.5
5 1 5
TI (44)
* From time of initial BCG treatment.
TABLE 4.
Tl Ta+Tis Ta Ta Tl+Tis Tl+Tis Tl+Tis
Initial T Stage
rence
Tl* Tl Ta+Tis Ta+* Tl+Tis Ta+Tis* Ta+Tis Tl* Tl+Tis Tis* Tl+Tis* Ta+Tis*
No Yes No Yes Yes Yes Yes Yes No No No No
Overall Surviva! (yrs.)
With Bladder Recurrence
Treatment of Progression
Noninvasive prostatic Ca involving urethra or ducts Stromal invasion Noninvasive prostatic Ca involving urethra or ducts Stromal invasion Noninvasive prostatic Ca involving urethra or ducts Stromal invasion
14
Radical cystectomy
PO
13
No
40 16
Radical cystectomy Radical cystectomy
P4
11
PO
12
Yes No
10 60
Radical cystectomy Radical cystectomy
Pis,N+
11
PO
12
No Yes
48
Dead of disease
No
Stromal invasion Noninvasive prostatic Ca involving urethra or ducts
108 84
Systemic chemotherapy Radical cystectomy Radical cystectomy
10 10
No No
Type of Progression
TABLE
Superficial Recur-
p Stage
Mos. to Progression
Initial T Stage Ta+Tis
Outcome of patients with progression in the prostate
5. Patients dead of cancer
Type of Progression
Mos. to Progression
T2+M+ M+ M+ T2+M+ M+t TisM+ M+
32 3 15 24 96 72 7 24 18 96 48 24
M+ T2+M+ T2M+ T4pM+ TlM+
Treatment of Progression
Overall Survival (yrs.)
Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy
3.4 1 1.8 2.3 9
Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy
8
2 3 2.5 9 5 2
* Indicates patients who might survive with earlier cystectomy.
t Metastases from renal pelvis tumor.
tasis even in the event of local progression. Against this possibility is the fact that all of these patients were entered into a randomized prospective trial in which 95% of the control population had progression (including 37% who had muscle infiltrating tumor) within 12 months. 3 This fact suggests that BCG may have conferred a durable protective effect against new tumor formation within regenerating bladder epithelium. In support of this observation is the decrease in tumor recurrence and progression rates in patients followed longer than 5 years, and the increasing detection with time of extravesical urothelial disease involving the upper urinary tracts and prostate. Topical BCG might not be expected to alter favorably neoplastic or preneoplastic events occurring at these sites. The current results are due most likely to a combination of factors, including an initial vigorous transurethral resection and the use of prophylactic or therapeutic BCG directed against minimal (Tis) disease, careful surveillance, repeat complete transurethral resection with adjunctive BCG as needed and early detection of tumor progression. While not preserving
P4 Pis
bladder function, salvage cystectomy contributed significantly to survival. An alternative explanation is that patients who had progression with new muscle infiltrating tumors after a history of recurrent superficial disease may have had a biologically more favorable tumor than is associated with de novo muscle invasion. The short interval (average 33 months) to metastatic progression and the fact that not all patients (6 of 12) had evident cancer in the bladder at the time of such metastasis suggest understaging of the primary tumor and/ or unrecognized metastasis at initial diagnosis and treatment. Alternatively, it must be recognized that in some cases superficial tumors, especially stage Tl cancers, may not be adequately controlled with transurethral resection and BCG leading to undetected local progression and metastasis. That may have been the case in the patient who had relapse after 8 years with infiltrating disease in the bladder and concomitant distant dissemination. Thus, this patient coupled with the other 6 who had metastases after the development of progression in the bladder (5) or prostate (1) suggests an overall probability of 11 % (7 of 61) that earlier cystectomy might have prevented a cancer death. If we optimistically assume that early radical cystectomy in all 61 patients would have salvaged all 11 (18%) who had metastases from the bladder, and take into account the 13 (21 %) who were saved with delayed cystectomy, there remains 37 patients (61 %) who would have lost the bladder unnecessarily. Therefore, the overall risk of cancer death as the result of conservative bladder sparing therapy in this high risk group was 18%. Against this risk must be weighed the cost of preserving bladder (and sexual) function, including frequent surveillance, multiple invasive diagnostic and therapeutic procedures, and risk of extravesical disease, which at least in cases involving the distal ureters and prostate probably would be prevented by cystectomy. Based upon our current data we believe the potential for long-term bladder preservation, progression-free and overall survival justifies the risk and cost of
B_ACILLUS CALlV!ETTE-GUERI:N IPsJ SUPERFICIAL BLAI)DER CAI'+JCER
BCG therapy and fr,,,"'''°r,t followup evaluations for most patients with recurrent superficial bladder cancer. REFERENCES
1. Lamm, D. L., Thor, D. E., Winters, W. D., Stogdill, V. D. and
Radwin, H. M.: ECG immunotherapy of bladder cancer: inhibition of tumor recurrence and associated immune responses. Cancer, 48: 82, 1981. 2. Pinsky, C. M., Camacho, F. J., Kerr, D., Geller, N. L., Klein, F. A., Herr, H. A., Whitmore, W. F., Jr. and Oettgen, H.F.: Intravesical administration of bacillus Calmette-Guerin in patients with recurrent superficial carcinoma of the urinary bladder: report of a prospective, randomized trial. Cancer Treat. Rep., 69: 47, 1985. 3. Herr, H. W., Laudone, V. P., Badalament, R. A., Oettgen, H. F., Sogani, P. C., Freedman, B. D., Melamed, M. R. and Whitmore, W. F., Jr.: Bacillus Calmette-Guerin therapy alters the progression of superficial bladder cancer. J. Clin. Oncol., 6: 1450, 1988.
EDITORIAL COMMENT The authors provide an additional 5 years of followup on their remarkable controlled comparison of surgery versus surgery plus BCG immunotherapy reported in 1988 (reference 3 in article). The original study demonstrated that BCG significantly delayed disease progression and decreased mortality. The current study includes 18 original control patients who received BCG after multiple recurrences and is presented as a historical review rather than a long-term controlled comparison. The benefit of therapy is difficult to estimate in historical series. Clearly, as the authors state, the population studied was comprised of patients at high risk for tumor recurrence and progression. For example, 72% of the patients had carcinoma in situ in addition to recurrent stage Ta or Tl transitionai eel! carcinoma. The aggressive nature of these tumors may also be reflected in the high incidence (18%) of
J.023
subsequent upper tract tur:.i'lors in this population. In a recent teview of upper tract tumors in patients undergoing cy:ste,ct,Jm,y for bladder cancer, only 2.4% had tumors during follovvup. However-, even if we accept that this is, indeed, a high risk population, how good is a 10year recurrence-free rate of 31%? How bad is a 41% incidence of progression? Perhaps the best historical comparison to use for the study is the treatment generally recognized as optimal, radical cystectomy. In the current study the actual 10-year survival rate was 75%. In a recent review by Malkowicz et al the 5-year actuarial smvival for patients with carcinoma in situ plus stage Tl tumor was 78%. 2 These data suggest that BCG immunotherapy does not decrease survival when compared with early cystectomy but confirmation of this impression would require a controlled clinical trial. Since the authors have not reported the extended followup of their controlled trial, I suspect that the reduction of progression and mortality that was significant at 5 years is not significant at 10 years. If true, the change in results could reflect the use of BCG in control patients who had not already undergone cystectomy. The use of BCG in these patients was clearly justified by humanitarian concerns. If the suspected loss of statistical significance did, in fact, occur as a reflection of the subsequent use of BCG in the control population, this change in management might strengthen rather than diminish the rationale for more extensive use of BCG. Hopefully, additional comparative data from this important study will be forthcoming.
Donald L. Lamm Department of Urology West Virginia University Morgantown, West Virginia 1. Malkowicz, S. B. and Skinner, D. G.: Development of upper tract carcinoma after cystectomy for bladder carcinoma. Urology, 36: 20, 1990. 2. Malkowicz, S. B., Nichols, P., Lieskovsky, G., Boyd, S, D., Huffman, J. and Skinner, D. G.: The role of radical cystectomy in the management of high grade superficial bladder cancer (PA, Pl, PIS and P2). J. Urol., 144: 641, 1990.
Vol. 147, 1020-1023, April 1992
Printed in U.S.A.
BACILLUS CALMETTE-GUERIN THERAPY FOR SUPERFICIAL BLADDER CANCER: A 10-YEAR FOLLOWUP HARRY W. HERR, DAVID D. WARTINGER, WILLIAM. R. FAIR AND HERBERT F. OETTGEN From the Urologic Service, Department of Surgery and Clinical Immunology Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, New York
ABSTRACT
We report the outcome of 61 patients with superficial bladder tumors who received bacillus --Galme-tte-Gu-e-r-in (B8G}the-rapy--and -were followed for at lea-st 10-ye-ars (range 10-to 13 years)-;--A
total of 19 patients (31 %) remains free of tumor and progression, 17 (28%) had interval superficial recurrences but no progression and 25 (41 %) had progression, first identified as muscle invasion in 12, prostatic involvement in 8 or metastasis in 5. Most tumors recurred or progressed within the first 5 years. Of the 61 patients 33 (54 %) are disease-free and with an intact bladder, 13 (21 %) are alive after cystectomy, 2 (3%) died of other causes, 1 (2%) is alive with metastasis and 12 (20%) died of metastatic urothelial cancer (11 from the bladder or prostate and 1 from an upper tract tumor). KEY WORDS:
bladder neoplasms; BCG vaccine; carcinoma, transitional cell
lntravesical bacillus Calmette-Guerin (BCG) has been shown in randomized clinical trials to alter the pattern of recurrence1· 2 and progression 3 of superficial bladder tumors. However, such studies do not provide absolute followup of more than 5 years and, therefore, they do not address the question of whether BCG simply delays or actually prevents disease progression. In other words, can any intravesical agent abrogate tumor progression and preserve the bladder, or will ultimate sacrifice of the bladder (and perhaps lives) be an inexorable consequence of the neoplastic diathesis? We report the outcome of high risk bladder cancer patients treated by transurethral resection and BCG for intervals of more than 10 years. MATERIAL AND METHODS
A total of 86 patients with recurrent superficial bladder cancer from 1978 to 1981 was entered into a previously reported prospective controlled clinical trial. 2 All patients had papillary noninvasive (stage Ta) or superficially invasive (stage Tl) tumors (grades 2 and 3), with or without associated carcinoma in situ (Tis). The initial objective was to determine the effect of BCG treatment on the recurrence of superficial bladder cancer after transurethral resection. Of the patients 43 were randomized to receive transurethral resection plus BCG and 43 underwent transurethral resection alone. Of the control (transurethral resection) patients 18 were crossed over to the BCG arm after 6 months to 1 year because of frequent tumor recurrences. These 61 ECG-treated patients have been followed for a minimum: of 10 years and are the focus of this report. BCG (Armand Frappier substrain) was administered intravesically (120 mg. in 50 ml. saline) and percutaneously (5 X 10 7 units). Treatment was begun 2 to 4 weeks after complete transurethral resection of overt papillary lesions and continued weekly for 6 weeks. BCG treatments were repeated intravesically at the same induction dosage and schedule. None of the patients received maintenance BCG. Patients were followed every 3 to 4 months thereafter with cystoscopy, biopsy or transurethral resection and urine cytology. After 5 years disease-free patients were generally evaluated at 6-month intervals. End points selected for analysis included number of patients with and without superficial (stages Ta, Tis and Tl) recurrences, number of patients with progression and the interval to Accepted for publication August 16, 1991. Supported by National Institutes of Health Grants CA-19267 and CA-74146.
progression, and survival from the initial BCG treatment. Tumor progression was defined as muscle invasion (stage T2+ ), metastasis (M+) or prostatic involvement (stage T4p). Prostatic disease included noninvasive urethral and/or ductal carcinoma and stromal invasion. RESULTS
Patient characteristics are listed in table 1. Of the 61 patients 34 (56%) had stage Ta and 27 (44%) had stage Tl tumors. A total of 44 patients (72%) had coexisting carcinoma in situ (Tis) proved by biopsy and positive urine cytology, and they received BCG therapeutically. The other 17 patients underwent resection of papillary lesions only, and they had negative posttransurethral resection cytology studies and received prophylactic BCG. Of the patients 41 received 1 course of BCG, and 20 received 2 (14 patients) or 3 (6 patients) 6-week treatments. The average interval between BCG treatments was 26 months. Sequential BCG courses (that is 12 consecutive weeks) were not given. A total of 19 patients (31 %) remained free of recurrence and progression for more than 10 years after transurethral resection and 1 course of BCG: 1 had stage Ta, 6 had stage Ta + Tis, 1 had stage Tl and 11 had stage Tl + Tis disease. Another 17 TABLE
1. Patient characteristics
No. pts. Median age (range) No. men/No. women Initial T stage (No. pts.): Ta Ta+ Tis Tl Tl+ Tis No. BCG treatments (No. pts.): 1 2 3
61 63 (34-77) 47/14 11 23 6
21 41 14 6
No. pts. without tumor progression (%):
No recurrences Recurrences Type of progression (% ): T2+
T4 (p) M+ Total No.(%) Followup (yrs.): Minimum Median (range)
1020
19 17
(31) (28)
12 8
(20) (13)
5
(8)
25
(41)
10 11.9 (10-13)
1021
BACILLUS CALMETTE-GUEIUN IN SUPERFICIAL BLADDER CANCER
patients (28%) had recunent superficial bladder tumors but no progression during the 10-year period. These patients received BCG initially for stage Ta tumors in 6, stage Ta + Tis in 5 and stage Tl + Tis in 6. The outcome of these 36 patients without progression is shown in table 2. The number of patients with positive cystoscopic evaluations showing recurrent tumor indicates that tumor recurred most often within the initial 5 years (the majority were seen within 2 years). Eight patients had recurrence(s) before but none after 5 years, 2 had no recurrences within 5 years only to suffer recurrent disease at 6 and 7 years, and 7 had sporadic recurrences during the entire 10+-year period. The trend was toward fewer recurrences among fewer patients after 5 and up to 10 years of followup. Recurrences were generally easily controlled by transurethral resection, which was supplemented in some patients with carcinoma in situ by additional BCG. Cystectomy was not done in any patient for recurrent superficial bladder tumors alone. Only 1 patient in this group died of metastatic renal pelvis cancer. There was no difference in the initial T stage among patients who did or did not have superficial tumor recurrence. Progression occurred in 25 patients (41 %): 12 (20%) with muscle invasion, 8 (13%) with prostatic involvement and 5 (18%) with distant metastasis (table 1). The outcome of patients with such progression is shown in table 3. The distribution in initial T stages (stage Ta in 5, stage Ta + Tis in 10, stage Tl in 4 and stage Tl + Tis in 6) did not differ from those in patients without progression. Tumor progressed most often within 5 years, although 3 patients had muscle infiltrating tumors at 6, 8 and 9 years, respectively. No tumor progressed after 10 years. Of the 25 patients 8 (32%) had superficial bladder recurrences at the time or within 6 months before disease progression was documented. The other 13 patients (52%) had de novo progression as specified that was not heralded by a pattern of prior recurrence. Of the 12 patients with stage T2+ tumors 4 had superficial lesions 3 to 6 months previously, and in 3 this was a stage Tl tumor. Radical cystectomy (transurethral resection in 1 patient) salvaged all but 1 patient with progression only in the bladder or prostate. None of the patients with metastasis (with or without local disease) survived beyond 5 years despite chemotherapy. A muscle infiltrating (stage T2+) tumor developed in 12 patients at an average of 33 months after BCG therapy. Four patients had metastases at the time or within 3 months of detection of the bladder cancer and 1 had metastases from an invasive bladder tumor after 8 years. The average interval to stage T2+ progression in this group was 26 months. Seven patients had no distant disease when the stage T2+ bladder tumor occurred. The average interval to invasion among these patients was 45 months. All patients remain disease-free after salvage therapy (table 3). Table 4 shows the outcome of 8 patients with progression in the prostate: prostatic transitional carcinoma was diagnosed within 5 years in 5, and 3 had disease at 5, 7 and 9 years, respectively. The presence or absence of concomitant bladder tumor recurrence was a poor predictor of subsequent prostatic disease, since the bladder was free of disease in 6 patients when the prostatic tumor was documented. One patient with stromal TABLE 2.
Superficial Bladder Re-
No. Pts.
No. Pts. With Recurrences Between
invasion died but the other 7 (5 with noninvasive ductal tumor and 2 with stromal invasion) were salvaged with cystoprostatectomy. Table 5 shows characteristics of the 12 patients who died of urothelial cancer. All 12 patients died of metastases emanating from the bladder (11) or upper tract (1). Five patients had metastases without an invasive bladder tumor, although 4 had had prior superficial recurrences. The other 7 patients had concomitant local and systemic progression, which in 5 was not heralded by a pattern of recurrent superficial tumors. If we assume that distant metastases appearing within 2 years of BCG therapy were from the original bladder tumor (4 patients), then 7 of these patients might have been saved if cystectomy had been performed initially. DISCUSSION
Assessing the impact of conservative management on tumor recurrence, progression and survival in patients with superficial bladder tumors is compromised by the often long natural history of the disease, the competing mortality in the age groups concerned and the lack of studies providing long-term followup. Can some patients be cured, or at least maintain a functioning bladder for more than 10 years? Are enough patients with progression salvaged to justify the initial conservative approach? Of the 61 patients followed for at least 10 and up to 13 years 33 (54%) have a functioning bladder (albeit half required local therapy for tumor recurrence), 13 (21 %) are disease-free after cystectomy, 2 (3%) are dead of unrelated causes, 1 (2%) is currently alive with metastatic disease and 12 (20%) died of metastatic bladder/prostate (11) or upper tract (1) urothelial cancer. Progression of bladder cancer was noted more often within the initial 5 years, suggesting that surveillance should be particularly intense during this interval even among patients without superficial recurrence. Tumor progression after 5 years and up to 10 years was sufficiently common to make lifelong surveillance for panurothelial cancer a necessity in such patients. A total of 11 patients (18%) had upper tract tumors from 1.5 to 9 years after BCG treatment: in 2 metastasis originated from a renal pelvis tumor, while 9 were salvaged with either distal ureterectomy or nephroureterectomy. These patients will be the subject of a separate report. The initial T stage and the presence or number of subsequent superficial bladder tumor recurrences did not correlate with progression. An average of 12 cystoscopy studies identified an average of 38 tumors among patients with superficial recurrences but no progression compared to 5 positive cystoscopy studies showing 14 superficial tumors among patients with progression. Furthermore, because so many of the patients (72%) had associated in situ carcinoma the impact of multifocal disease upon recurrence and progression rates cannot be well evaluated. Why did these high risk patients have a progression rate of less than 50% during 10 years, a high rate of salvage when nonmetastatic progression occurred and a favorable overall survival? An explanation might be selection of patients with tumors having a long natural history and a low rate of metas-
Outcome of patients without progression Av. No. Pos. Cystoscopy Studies (Av. No. Tumors) Between
currences
0-5 Yrs.
Neg. Pos. Totals(%)
19 17
36 (59)
15
5-10 Yrs. 9
0-5 Yrs. 5.7 (20)
* Cerebrovascular accident at 9 years and myocardial infarction at 6 years.
t Metastasis from upper tract tumor.
5-10 Yrs. 3.2 (7)
No. Pts. Disease-Free at More Than 10 Yrs.
No. Pts. Dead of Dis-
No. Pts. Dead of Other Causes
No. Pts. Alive With Disease
ease
16 16
0
32 (52)
I
1t
2* 0
2
1t 0
I
1022
HERR AND ASSOCIATES TABLE 3.
No. Pts. ( %)
Type of Progression
T2+ (muscle invasion) T4p (prostatic involvement) M+ (metastasis) Totals(%)
Outcome of patients with progression
No. Pts. With Progression Between
Av. Mos. to Progression
0-5 Yrs.
5-10 Yrs.
No. Pts. With Bladder Recurrences 0-6 Mos. Before Progression
12 (20)
33
10
2
4
8 (13)
47 20
5 5
3
2 2
20
5
5 (8)
25 (41)
0
No. Pts. With No Evidence of Disease
Treatment of Progression Radical cystectomy Transurethral resection Systemic chemotherapy Radical cystectomy Systemic chemotherapy
8 (32)
6 1 0 7 0 14(56)
Av. Overall Survival (yrs.)*
No. Pts. Dead of Disease
11
10.8 3.5
5 1 5
TI (44)
* From time of initial BCG treatment.
TABLE 4.
Tl Ta+Tis Ta Ta Tl+Tis Tl+Tis Tl+Tis
Initial T Stage
rence
Tl* Tl Ta+Tis Ta+* Tl+Tis Ta+Tis* Ta+Tis Tl* Tl+Tis Tis* Tl+Tis* Ta+Tis*
No Yes No Yes Yes Yes Yes Yes No No No No
Overall Surviva! (yrs.)
With Bladder Recurrence
Treatment of Progression
Noninvasive prostatic Ca involving urethra or ducts Stromal invasion Noninvasive prostatic Ca involving urethra or ducts Stromal invasion Noninvasive prostatic Ca involving urethra or ducts Stromal invasion
14
Radical cystectomy
PO
13
No
40 16
Radical cystectomy Radical cystectomy
P4
11
PO
12
Yes No
10 60
Radical cystectomy Radical cystectomy
Pis,N+
11
PO
12
No Yes
48
Dead of disease
No
Stromal invasion Noninvasive prostatic Ca involving urethra or ducts
108 84
Systemic chemotherapy Radical cystectomy Radical cystectomy
10 10
No No
Type of Progression
TABLE
Superficial Recur-
p Stage
Mos. to Progression
Initial T Stage Ta+Tis
Outcome of patients with progression in the prostate
5. Patients dead of cancer
Type of Progression
Mos. to Progression
T2+M+ M+ M+ T2+M+ M+t TisM+ M+
32 3 15 24 96 72 7 24 18 96 48 24
M+ T2+M+ T2M+ T4pM+ TlM+
Treatment of Progression
Overall Survival (yrs.)
Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy
3.4 1 1.8 2.3 9
Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy Systemic chemotherapy
8
2 3 2.5 9 5 2
* Indicates patients who might survive with earlier cystectomy.
t Metastases from renal pelvis tumor.
tasis even in the event of local progression. Against this possibility is the fact that all of these patients were entered into a randomized prospective trial in which 95% of the control population had progression (including 37% who had muscle infiltrating tumor) within 12 months. 3 This fact suggests that BCG may have conferred a durable protective effect against new tumor formation within regenerating bladder epithelium. In support of this observation is the decrease in tumor recurrence and progression rates in patients followed longer than 5 years, and the increasing detection with time of extravesical urothelial disease involving the upper urinary tracts and prostate. Topical BCG might not be expected to alter favorably neoplastic or preneoplastic events occurring at these sites. The current results are due most likely to a combination of factors, including an initial vigorous transurethral resection and the use of prophylactic or therapeutic BCG directed against minimal (Tis) disease, careful surveillance, repeat complete transurethral resection with adjunctive BCG as needed and early detection of tumor progression. While not preserving
P4 Pis
bladder function, salvage cystectomy contributed significantly to survival. An alternative explanation is that patients who had progression with new muscle infiltrating tumors after a history of recurrent superficial disease may have had a biologically more favorable tumor than is associated with de novo muscle invasion. The short interval (average 33 months) to metastatic progression and the fact that not all patients (6 of 12) had evident cancer in the bladder at the time of such metastasis suggest understaging of the primary tumor and/ or unrecognized metastasis at initial diagnosis and treatment. Alternatively, it must be recognized that in some cases superficial tumors, especially stage Tl cancers, may not be adequately controlled with transurethral resection and BCG leading to undetected local progression and metastasis. That may have been the case in the patient who had relapse after 8 years with infiltrating disease in the bladder and concomitant distant dissemination. Thus, this patient coupled with the other 6 who had metastases after the development of progression in the bladder (5) or prostate (1) suggests an overall probability of 11 % (7 of 61) that earlier cystectomy might have prevented a cancer death. If we optimistically assume that early radical cystectomy in all 61 patients would have salvaged all 11 (18%) who had metastases from the bladder, and take into account the 13 (21 %) who were saved with delayed cystectomy, there remains 37 patients (61 %) who would have lost the bladder unnecessarily. Therefore, the overall risk of cancer death as the result of conservative bladder sparing therapy in this high risk group was 18%. Against this risk must be weighed the cost of preserving bladder (and sexual) function, including frequent surveillance, multiple invasive diagnostic and therapeutic procedures, and risk of extravesical disease, which at least in cases involving the distal ureters and prostate probably would be prevented by cystectomy. Based upon our current data we believe the potential for long-term bladder preservation, progression-free and overall survival justifies the risk and cost of
B_ACILLUS CALlV!ETTE-GUERI:N IPsJ SUPERFICIAL BLAI)DER CAI'+JCER
BCG therapy and fr,,,"'''°r,t followup evaluations for most patients with recurrent superficial bladder cancer. REFERENCES
1. Lamm, D. L., Thor, D. E., Winters, W. D., Stogdill, V. D. and
Radwin, H. M.: ECG immunotherapy of bladder cancer: inhibition of tumor recurrence and associated immune responses. Cancer, 48: 82, 1981. 2. Pinsky, C. M., Camacho, F. J., Kerr, D., Geller, N. L., Klein, F. A., Herr, H. A., Whitmore, W. F., Jr. and Oettgen, H.F.: Intravesical administration of bacillus Calmette-Guerin in patients with recurrent superficial carcinoma of the urinary bladder: report of a prospective, randomized trial. Cancer Treat. Rep., 69: 47, 1985. 3. Herr, H. W., Laudone, V. P., Badalament, R. A., Oettgen, H. F., Sogani, P. C., Freedman, B. D., Melamed, M. R. and Whitmore, W. F., Jr.: Bacillus Calmette-Guerin therapy alters the progression of superficial bladder cancer. J. Clin. Oncol., 6: 1450, 1988.
EDITORIAL COMMENT The authors provide an additional 5 years of followup on their remarkable controlled comparison of surgery versus surgery plus BCG immunotherapy reported in 1988 (reference 3 in article). The original study demonstrated that BCG significantly delayed disease progression and decreased mortality. The current study includes 18 original control patients who received BCG after multiple recurrences and is presented as a historical review rather than a long-term controlled comparison. The benefit of therapy is difficult to estimate in historical series. Clearly, as the authors state, the population studied was comprised of patients at high risk for tumor recurrence and progression. For example, 72% of the patients had carcinoma in situ in addition to recurrent stage Ta or Tl transitionai eel! carcinoma. The aggressive nature of these tumors may also be reflected in the high incidence (18%) of
J.023
subsequent upper tract tur:.i'lors in this population. In a recent teview of upper tract tumors in patients undergoing cy:ste,ct,Jm,y for bladder cancer, only 2.4% had tumors during follovvup. However-, even if we accept that this is, indeed, a high risk population, how good is a 10year recurrence-free rate of 31%? How bad is a 41% incidence of progression? Perhaps the best historical comparison to use for the study is the treatment generally recognized as optimal, radical cystectomy. In the current study the actual 10-year survival rate was 75%. In a recent review by Malkowicz et al the 5-year actuarial smvival for patients with carcinoma in situ plus stage Tl tumor was 78%. 2 These data suggest that BCG immunotherapy does not decrease survival when compared with early cystectomy but confirmation of this impression would require a controlled clinical trial. Since the authors have not reported the extended followup of their controlled trial, I suspect that the reduction of progression and mortality that was significant at 5 years is not significant at 10 years. If true, the change in results could reflect the use of BCG in control patients who had not already undergone cystectomy. The use of BCG in these patients was clearly justified by humanitarian concerns. If the suspected loss of statistical significance did, in fact, occur as a reflection of the subsequent use of BCG in the control population, this change in management might strengthen rather than diminish the rationale for more extensive use of BCG. Hopefully, additional comparative data from this important study will be forthcoming.
Donald L. Lamm Department of Urology West Virginia University Morgantown, West Virginia 1. Malkowicz, S. B. and Skinner, D. G.: Development of upper tract carcinoma after cystectomy for bladder carcinoma. Urology, 36: 20, 1990. 2. Malkowicz, S. B., Nichols, P., Lieskovsky, G., Boyd, S, D., Huffman, J. and Skinner, D. G.: The role of radical cystectomy in the management of high grade superficial bladder cancer (PA, Pl, PIS and P2). J. Urol., 144: 641, 1990.
