Vaughn 2005
Vaughn 2005
Vaughn 2005
This year, approximately 63,210 Americans from these data that radical cystectomy and pelvic
will be diagnosed with bladder cancer, and an lymph node dissection is an excellent treatment for
estimated 13,180 Americans will die of this disease organ-confined disease, many patients with extra-
[1]. Muscle-invasive bladder cancer is a highly vesical or lymph node–positive bladder cancer will
lethal disease with a mortality rate of 70% to 90% develop recurrent disease, often with distant me-
over 2 to 3 years if left untreated. Radical tastases, and will ultimately die of their disease.
cystectomy and bilateral pelvic lymph node dis- Given the lethality of muscle-invasive bladder
section can have a significant impact on survival, cancer, there is a definite need for effective systemic
but outcomes are significantly determined by chemotherapy.
pathologic stage [2].
In a large series of 1054 patients treated with
radical cystectomy and bilateral pelvic lymph node Chemotherapy for advanced urothelial cancer
dissection, investigators at the University of Bladder cancer is a chemotherapy-sensitive
Southern California’s Norris Comprehensive Can- neoplasm. Significant single-agent tumor response
cer Center demonstrated that T and N stage were rates have been demonstrated with several single
important predictors of prognosis [3]. These results agents. Cisplatin-based chemotherapy regimens
are shown in Table 1 (the 1987 Tumor-Node- have been most extensively investigated. In the
Metastasis (TNM) system was used in this study). 1980s, investigators at Memorial Sloan-Kettering
For patients who have organ-confined muscle- Cancer Center (MSKCC) developed the metho-
invasive lymph node–negative disease, the 5- and trexate, vinblastine, doxorubicin, and cisplatin
10-year survival rates were 89% and 78%, re- (MVAC) regimen. Early phase II investigations
spectively, for pT2 tumors and 87% and 76%, of MVAC demonstrated tumor response rates in
respectively, for pT3a tumors. Patients with extra- the 70% to 80% range [4]. Subsequently, ran-
vesical lymph node–negative disease had signifi- domized trials have demonstrated that MVAC is
cantly worse 5- and 10-year survival: 62% and superior to single-agent cisplatin and to CISCA
61%, respectively, for pT3b tumors and 50% and (cyclophosphamide, doxorubicin and cisplatin)
45%, respectively, for pT4 tumors. The 5- and 10- [5,6]. The Intergroup evaluated 246 patients with
year survival rates for lymph node–positive disease advanced urothelial cancer who were randomized
were 35% and 34%, respectively. While it is clear to single agent cisplatin or MVAC. The MVAC
arm was associated with a higher tumor response
* Corresponding author. rate and improved overall survival [5]. Investiga-
E-mail address: [email protected] tors at M.D. Anderson Cancer Center random-
(S.B. Malkowicz). ized patients with advanced bladder cancer to
0094-0143/05/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ucl.2005.03.001 urologic.theclinics.com
232 VAUGHN & MALKOWICZ
controlled trials of neoadjuvant chemotherapy for or cystectomy alone was significant, with 85%
invasive bladder cancer that included a total of and 82% of patients alive at 5 years, respectively.
3315 eligible randomized patients. In addition, Second, patients enrolled in the trial were strati-
data from previously published meta-analyses fied by initial tumor stage (T2 vs T3 or T4a) and
were used. Given inconsistencies and incomplete age (!65 years vs R65 years). Neither initial
data from some trials, a total of 11 trials with stage nor patient age had a significant effect on
a total of 2605 patients were ultimately examined. survival. Finally,the administration of neoadju-
The pooled hazard ratio from these 11 trials was vant MVAC did not prevent patients from un-
0.90 (95% CI 0.82–0.99; P = .02). This result was dergoing radical cystectomy, nor was it associated
consistent with a 10% decrease in the risk of death with more surgery-related complications. In fact,
or an absolute survival advantage of 5.0% (95% unlike trials of the MVAC combination in meta-
CI 0.0.5–9.0) from 50% to 55%. When the 8 trials static bladder cancer, there were no chemother-
that used cisplatin-based combination chemother- apy-related deaths. It is not clear whether this was
apy were examined, the pooled hazard ratio was related to increased familiarity of the regimen by
0.87 (95% CI 0.78–0.96; P = .006). This hazard the treating oncologists, better performance status
ratio was consistent with an absolute survival and health of the patients being treated, or simply
advantage of 6.5% (95% CI 2%–11%) from chance.
50 to 56.5%. Interestingly, when the 3 trials
that used single-agent cisplatin were examined
(a total of 337 patients), the survival benefit was
The importance of surgical factors in examining the
not demonstrated. The meta-analysis of the ran-
results of neoadjuvant chemotherapy in bladder
domized controlled trials of neoadjuvant chemo-
cancer
therapy in bladder cancer did not show statistical
heterogeneity, and the authors report that An important supplemental analysis of the
the hazard ratios are of a ‘‘fixed effect model.’’ INT-0080 trial demonstrated the importance of
Overall, the rate of chemotherapy-related mor- considering surgical factors when evaluating large
tality was 1.1%. Therefore, this meta-analysis cystectomy chemotherapy trials [21]. Patients
demonstrates that cisplatin-based chemother- who had complete surgical records and who
apy improves survival in muscle-invasive bladder underwent cystectomy according to protocol
cancer. were analyzed (n = 268). The cystectomies
were perfomed by 106 surgeons at 109 institu-
tions. Half of the patients received neoadjuvant
chemotherapy.
Lessons from one trial: INT-0080
Five-year postcystectomy survival rates were
The United States Intergroup (INT)-0080 trial 54%, and local recurrence rates were 15%. In
[20] was reported in 2003. In INT-0080, 317 a multivariate model adjusted for MVAC admin-
patients with T2-T4a bladder cancer were ran- istration, age, pathologic stage, and node status
domized to 3 cycles of neoadjuvant MVAC demonstrated that surgical variables associated
followed by cystectomy versus cystectomy alone. with longer postcystectomy survival included
Using intention-to-treat analysis, the median negative surgical margins (Hazard Ratio (HR)
survival of patients in the cystectomy alone arm 0.37; P = .0007) and 10 or more nodes removed
was 46 months versus 77 months in the neo- (HR 0.51; P = .0001). These associations did not
adjuvant chemotherapy plus cystectomy arm vary by treatment arms (P O .21). Predictors of
(P = .06 by two-sided stratified log-rank test). local recurrence in adjusted multivariate models
There are several issues of interest that are were positive margins (odds ratio 11.2;
demonstrated in the trial. First, complete down- P = .0001) and fewer than 10 nodes removed
staging to a no residual disease status at cystec- (odds ratio 5.1; P = .002). The conclusion of this
tomy had a significant impact on survival. More analysis was that surgical factors can influence
patients in the neoadjuvant chemotherapy plus cystectomy outcomes apart from pathologic fac-
cystectomy group demonstrated no residual can- tors or neoadjuvant treatment status. Thus the
cer (that is, pT0 disease status) than in the extent and uniformity of cystectomy and bilateral
cystectomy alone arm (38% vs 15%; P ! .001). pelvic lymph node dissection is a factor to be
The survival benefit of a pT0 disease status in considered in trials involving the perioperative use
patients treated with neoadjuvant chemotherapy of chemotherapy.
NEOADJUVANT CHEMOTHERAPY IN BLADDER CANCER 235
Neoadjuvant chemotherapy for bladder cancer be assumed that all regimens are the same as
in perspective MVAC. However, the ongoing EORTC/SWOG
adjuvant trial incorporates MVAC, high-dose
Selected randomized controlled cisplatin-based
intensity MVAC, and gemcitabine/cisplatin in
combination chemotherapy trials demonstrate
the adjuvant chemotherapy arm.
survival benefit, which is supported by a recently
Currently the literature concerning periop-
published meta-analysis [19]. These data on neo-
erative chemotherapydbe it neoadjuvant or
adjuvant chemotherapy do not irrefutably estab-
adjuvantdwhen viewed as a whole provides
lish this sequence of therapy as the new standard
strong support for the use of perioperative che-
of care for patients with muscle-invasive bladder
motherapy in patients with poor-risk bladder
cancer. Randomized trials of adjuvant chemother-
cancer who are to undergo cystectomy. At the
apy in bladder cancer have also been performed
Abramson Cancer Center of the University of
[22]. Although these trials suffer in general from
Pennsylvania, our general approach is to discuss
being underpowered and problematic with respect
the available information on perioperative ther-
to trial methodology, two trials suggest survival
apy with cystectomy candidates. Those with
benefit [23,24]. Presently, the EORTC and the
bulky, local disease are encouraged to consider
SWOG are conducting a randomized trial of
neoadjuvant therapy, while individuals with low-
adjuvant chemotherapy in high-risk bladder can-
volume/stage disease usually undergo cystectomy.
cer patients. This trial randomizes patients with
The use of adjuvant therapy is based on patho-
pT3 or pT4 or node-positive bladder cancer to
logical findings and the patient’s overall general
either early postoperative chemotherapy versus
health status. If a patient has pT3 or pT4 or node-
chemotherapy deferred until the time of disease
positive disease, and can tolerate chemotherapy,
progression. The chemotherapy regimens that are
we will offer the patient four cycles of adjuvant
being evaluated include MVAC, high-dose in-
cisplatin-based chemotherapy (either MVAC or
tensity MVAC, and gemcitabine/cisplatin. With
the gemcitabine/cisplatin, two regimens with sim-
an overall accrual goal of 1344 patients, the trial is
ilar efficacy in the advanced disease setting).
sufficiently powered to demonstrate a 5% survival
However, we make clear in our discussion with
difference at 5 years.
the patient that our recommendation is based on
Is there an advantage to preoperative neo-
supportive but not definitive evidence, and that it
adjuvant chemotherapy as opposed to postoper-
is the best we can offer until more definitive data
ative adjuvant chemotherapy in bladder cancer?
are available (eg, the ongoing EORTC/SWOG
In an important study examining this issue in
trial).
bladder cancer, Millikan and colleagues [25]
randomized 140 patients with muscle-invasive
disease to two cycles of neoadjuvant MVAC
Summary
chemotherapy followed by cystectomy plus pelvic
lymph node dissection, followed by three more Neoadjuvant chemotherapy has been exten-
cycles of adjuvant MVAC; or to cystectomy sively investigated in muscle-invasive bladder
followed by five cycles of adjuvant MVAC cancer. When taken together, the randomized
chemotherapy. Of note, at a median follow-up controlled trials of neoadjuvant cisplatin-based
of almost 6 years, 58% of all patients were alive combination chemotherapy demonstrate an im-
and disease-free. However, there was no statisti- proved survival over cystectomy alone. In addi-
cally significant difference in survival between the tion, neoadjuvant chemotherapy can result in
two arms. This is consistent with trials in other downstaging of primary tumors. As noted,
chemotherapy-sensitive disease sites (eg, the a pT0 disease status at cystectomy is associated
breast) that have not proven a survival advantage with a significant improvement in survival. A
to neoadjuvant chemotherapy over adjuvant randomized controlled trial comparing neoad-
treatment [26]. juavnt to adjuvant cisplatin-based chemotherapy
It is with caution that we must point out that shows that neither approach is superior. Finally,
most of the newer non–cisplatin-based chemo- the ongoing EORTC/SWOG adjuvant chemo-
therapy regimens, and even the gemcitabine/ therapy trial, when completed, should add impor-
cisplatin regimen, which are currently being used tantly to the literature concerning the role of
in metastatic disease, have not been evaluated in systemic chemotherapy in muscle-invasive bladder
the adjuvant or neoadjuvant setting. It should not cancer.
236 VAUGHN & MALKOWICZ
Results of a controlled prospective study. J Urol operative and postoperative M-VAC. J Clin Oncol
1992;148:302–7. 2001;19:4005–13.
[25] Millikan R, Dinney C, Swanson D, et al. Integrated [26] Fisher B, Bryant J, Wolmark N, et al. Effect of pre-
therapy for locally advanced bladder cancer: final operative chemotherapy on the outcome of women
report of a randomized trial of cystectomy plus with operable breast cancer. J Clin Oncol 1998;16:
adjuvant M-VAC versus cystectomy with both pre- 2672–85.