|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Integrated Therapy for Locally Advanced Bladder Cancer: Final Report of a Randomized Trial of Cystectomy Plus Adjuvant M-VAC Versus Cystectomy With Both Preoperative and Postoperative M-VACByFrom the Center for Genitourinary Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX. Address reprint requests to Randall E. Millikan, PhD, MD, M.D. Anderson Cancer Center, Box 013, 1515 Holcombe Blvd, Houston, TX 77030; email: rmillika{at}notes.mdacc.tmc.edu
PURPOSE: We conducted a phase III trial to investigate the timing of chemotherapy with respect to surgery for patients with resectable but high-risk urothelial cancer. The trial was also designed to evaluate the accuracy of clinical staging in patients with locally advanced cancer and the prognostic significance of chemotherapy-induced downstaging. PATIENTS AND METHODS: A total of 140 uniformly evaluated patients with locally advanced urothelial cancer were studied. Planned treatment was five cycles of chemotherapy (M-VAC: methotrexate, vinblastine, doxorubicin, and cisplatin) plus radical cystectomy and pelvic lymph node dissection. Patients were randomly assigned to receive either two courses of neoadjuvant M-VAC followed by surgery plus three additional cycles of chemotherapy, or, alternatively, to have initial cystectomy followed by five cycles of adjuvant chemotherapy. RESULTS: There were no significant differences in outcome between the two groups. By intent-to-treat, 81 patients (58%) remain disease-free, with median follow-up of 6.8 years. We confirmed a high rate of clinical understaging in this cohort, especially among patients showing lymphovascular invasion on biopsy. Patients with no residual muscle-invasive disease at cystectomy after neoadjuvant chemotherapy were likely to be cured. CONCLUSION: These results lend further support to the impression from small randomized trials that, in a high-risk cohort, there is an improved cure fraction by the combination of multiagent chemotherapy and surgery, although we found no preferred sequence. Importantly, it is possible to select appropriate patients for such therapy on the basis of clinical staging information. These results establish a benchmark of outcome for this cohort.
ALTHOUGH CYSTECTOMY remains the standard of care for muscle-invasive bladder cancer in the United States, local or distant failure continues to be distressingly common, reported in the range of 15% to 40% even for patients with disease that is pathologically confined within the muscle proper of the bladder wall (ie, stage pT2).1,2 With increasing T stage, and in particular for patients with clinically extravesicular cancer, the situation is much worse. Urothelial cancers that by clinical criteria are grossly beyond the confines of the bladder (ie, cT3b), invade adjacent organs (cT4a), or show lymphovascular invasion on the cystoscopic biopsy,3 are typically associated with poor surgical cure rates of 20% to 30%.4 These high-risk patients, who have a low expectation of cure from single-modality therapy with surgery only, are the subject of this report. In the face of the relatively poor prospect for cure with surgery only in this cohort, many investigators have reported various combinations of chemotherapy with surgery in an attempt to improve outcome for high-risk patients. In a retrospective analysis, Logothetis et al5 found that patients with pathologically extravesical disease treated with surgery and chemotherapy fared better than historical controls treated by cystectomy alone. The hypothesis raised by this retrospective review has subsequently been evaluated in several randomized trials. In the much-discussed study of Skinner et al,6 high-risk patients were randomized to either observation or adjuvant chemotherapy that usually consisted of cyclophosphamide, doxorubicin, and cisplatin. Notwithstanding that this was a small study in which many patients did not get the therapy to which they were assigned and that several different chemotherapy regimens were used, the results showed markedly more relapses in the surgery only group. Despite the small numbers, the results were statistically significant by intent-to-treat. Stöckle et al7 also reported a randomized trial of adjuvant chemotherapy versus observation. Only 83 patients were enrolled onto the study, which was stopped for ethical considerations. The mature results provided compelling evidence that adjuvant therapy applied to patients with pathologic T3b, pathologic T4, or pathologically node-positive (pN+) disease did indeed improve disease-free survival and cure fraction. More recently, Freiha et al8 reported a third small randomized trial. As in Germany, this trial was stopped early. Fifty-four patients were enrolled (only 50 were included in the published analysis), and patients with pathologic stage T3b, T4 or node-positive disease seemed to benefit from adjuvant chemotherapy. At the time of publication, 20 of 25 patients had experienced relapse in the surgery-only arm versus 13 of 25 in the combination arm. However, despite this difference, overall survival was not statistically different in this report, although the modest power of this small study and the effect of cross-over are clearly important considerations in interpreting the results of this trial. Data from a large randomized European trial of neoadjuvant therapy were recently reported.9 In this study of 976 patients with cT2 to cT4a disease, three cycles of cisplatin, methotrexate, and vinblastine were administered before definitive local therapy (which was left to investigator preference and was approximately evenly split between cystectomy and radical radiotherapy). The specifics of pretherapy examination under anesthesia (EUA) were not reported, but roughly two thirds of the patients were classified as T3 or higher. The results of the trial showed a trend in favor of the chemotherapy arm, with an observed all-cause median survival of 37.5 months in the local therapy-only group, compared with 44 months in the chemotherapy plus local therapy group. The 3-year survival rates were 50% and 55%, respectively (P = .075). Although we support the authors’ conclusion that neoadjuvant therapy cannot be recommended for patients with muscle-invasive bladder cancer, the study as reported sheds relatively little light on the contribution of chemotherapy for patients with truly locally advanced disease, such as those reported here. Results from an intergroup trial (primarily conducted by the Southwest Oncology Group) are expected this year. However, this trial also accrued a large number of patients with merely muscle-invasive disease, and thus only a subset of this experience will apply to patients with clinically apparent extravesical tumor extension. Each of these trials are easily criticized, and undoubtedly the magnitude of the benefit from chemotherapy in this context remains poorly defined. Nevertheless, in our view, the aggregate data from these trials do establish that there is a survival benefit and an improved cure fraction when adjuvant chemotherapy is applied to the subset with high-risk features. By contrast, we are not aware of any data that suggest a benefit from adding systemic chemotherapy to a cohort with pathologically organ-confined, node-negative urothelial cancer. In 1986, the departments of Genitourinary Medical Oncology and Urologic Surgery at the University of Texas M.D. Anderson Cancer Center decided to include chemotherapy as part of the routine care of patients with high-risk, locally advanced bladder cancer. The emerging data on the contribution of chemotherapy seemed compelling to us, and large, confirmatory trials were already being planned by the cooperative groups. Thus, it was decided to study the timing of chemotherapy with respect to surgery in a trial in which all patients would have the same overall burden of therapy: namely, radical cystectomy and five cycles of methotrexate, vinblastine, doxorubicin, and cisplatin (M-VAC). In addition, the trial was designed to investigate the accuracy of clinical staging in this patient population and, further, to specifically examine the impact of losing pathologic staging information because this information is necessarily lost in the context of neoadjuvant chemotherapy. A final objective was to assess the prognostic significance of pathologic downstaging as a result of neoadjuvant chemotherapy. Preliminary analyses of both medical and surgical aspects of this trial, which report on the first 102 patients, have been published.10,11 Here, we report the final results of this trial, which ultimately involved 140 patients uniformly staged and treated over a study period of 14 years.
Patients Eligible patients had high-risk but resectable urothelial cancer. Specifically, patients had either invasive disease that demonstrated lymphovascular invasion on a transurethral biopsy (TURBx), clinically extravesical disease as demonstrated by a three-dimensional mass on EUA, or involvement of adjacent organs, ie, cT4a disease, as shown by stromal invasion of the prostate (defined by pathologic analysis of prostatic biopsies) or direct extension into the vaginal wall. Patients with only induration, (ie, a two-dimensional mass on EUA), were not considered high-risk and were not eligible. Patients with fixation of the bladder on EUA (ie, cT4b) were not considered resectable and were also excluded. All patients had a chest x-ray and computerized tomography (CT) of the abdomen and pelvis to eliminate patients with radiographic evidence of regional adenopathy or distant metastatic disease. There was a low threshold to biopsy any suspicious findings under CT guidance, and patients found to have nodal involvement were not eligible. Patients had adequate physiologic reserve as shown by a left ventricular ejection fraction of at least 40% (by either echocardiography or nuclear medicine technique), and an estimated creatinine clearance of at least 40 mL/min (by Cockcroft-Gault method). Of note, the renal function requirement could be achieved by means of nephrostomy placement, which many patients required. Patients had an absolute neutrophil count of at least 2,000 cells/µL and a platelet count of at least 100,000/µL. Patients had a Zubrod performance status of 2 or better. No previous systemic chemotherapy was allowed, but patients could have any prior intravesical therapy. All patients provided written informed consent to participate in this clinical trial, which was approved by the institutional review board of the University of Texas M.D. Anderson Cancer Center.
Chemotherapy
Clinical Evaluation and Operative Treatment All patients had cancers for which the appropriate surgical management was considered to be radical cystectomy and pelvic lymphadenectomy, performed as previously described.11 Most patients had an ileal conduit. In general, this patient group was not thought to be ideal for orthotopic reconstruction, but some neobladders were constructed in particular cases.
Statistical Considerations
Patient Characteristics Between November 1986 and July 1998, 140 patients were registered. Baseline characteristics of these patients are listed in Table 1. In all, 86 patients had pure TCC, and 54 (39%) had at least one other histologic component. As expected, the patients were predominantly male. The median age at registration was 66 years; 46 patients (32%) were at least 70 years of age at registration. Of note, 56 (40%) had obstructive uropathy and required a nephrostomy tube at study entry, a telling indicator of the locally advanced nature of this cohort. Although principally targeted to patients with clinically extravascular bladder cancer, the protocol did allow the participation of patients with upper-tract primaries if they were high-grade and large enough to produce an obvious tumor mass on CT scan. Given the high rate of clinical understaging in such patients, it was felt that these patients also met the spirit of the study, which was to offer patients with a relatively low chance of surgical cure combined-modality therapy. Two patients with upper-tract primaries were enrolled onto the study.
Staging Among the 70 patients assigned initial surgery, 66 had primary bladder cancers and were pathologically staged (two patients had upper-tract primary cancers, one refused surgery, and one had incomplete surgical staging information). Of these 66, there was exact correspondence of clinical and pathologic stage in only 40% of cases. For example, of the 23 patients with clinically organ-confined disease (all of whom had lymphovascular invasion on the TURBx to be eligible), only nine patients (39%) were pathologically organ-confined. By contrast, of 39 patients clinically staged as extravesicular (cT3b) and assigned initial surgery, only five (13%) had organ-confined disease at cystectomy. Thus, clinical understaging was far more common than overstaging, and when viewed from the perspective of differences large enough to alter therapy, the correlation of clinical and pathologic stage was quite good. Overall, there were 15 patients (23%) with organ-confined disease who might have been managed differently on the basis of pathologic staging information (ie, may not have been viewed as high-risk and, thus, may not have been offered adjuvant chemotherapy). Of these 15, there were four who experienced treatment failure: one patient died of a perioperative myocardial infarction, and three patients relapsed. As shown in Figure 1A, this failure rate was not different from that of patients with node-negative, pathologically extravesicular cancers. That prognosis was not related to pathologic pT-stage was somewhat unexpected, and although this could be an artifact of small numbers, it suggests to us that chemotherapy really does have an effect, especially because pathologically non–organ-confined bladder cancers are generally not reported to be cured 70% of the time with surgery only.
As noted above and as others have reported,3,14,15 it is worth emphasizing that the presence of lymphovascular invasion on the TURBx was strongly predictive of higher pathologic stage. Among patients undergoing immediate surgery, there were 23 with clinically organ-confined disease but also with a finding of lymphovascular invasion on the preoperative cystoscopic biopsy. Fourteen (61%) of these 23 were upstaged to either extravesical extension (eight patients, 36%) or pN+ (six patients, 27%). Likewise, understaging with respect to nodal status was common in patients with cT3b disease by preoperative evaluation: 16 (42%) of 38 had occult nodal involvement discovered at cystectomy. Among those with cT4a disease, two (40%) of five were pN+. In all, 24 (36%) of 66 patients had occult nodal metastases despite aggressive attempts to identify and exclude such patients before surgery.
Response and Survival
Cause-specific survival by pathologic stage for the 66 patients with bladder cancer who were treated with initial surgery is shown in Fig 1A. Overall, of the 70 patients assigned initial surgery, 42 (60%) remain free of disease. Once again, in the setting of chemotherapy, the influence of pathologic T stage was not statistically significant, although the difference between pN- versus pN+ was highly significant, even in the face of small numbers (P = .011). Of the 70 patients assigned to initial chemotherapy, 39 (56%) remain free of disease; 63 of these 70 patients went on to have a cystectomy. After two cycles of M-VAC, 25 (40%) of 63 had no evidence of residual muscle-invasive disease in the resected specimen. Only three (12%) of these 25 patients experienced relapse, although there were two additional patients who experienced treatment failure in this group, one patient who suffered a fatal postoperative myocardial infarction and another who succumbed to complications of chemotherapy in the third cycle of M-VAC. On the other end of the spectrum of response, those with pathologic involvement of pelvic nodes despite two cycles of M-VAC fared especially poorly. Of 14 such patients, 12 (86%) subsequently experienced relapse and died of metastatic cancer (pN+ v pN-, P < .0001). Cause-specific survival by pathologic stage after two cycles of M-VAC is shown in Fig 1B. Of note, all patients with pathologically extravesical extension (ie, pT3 disease) also had nodal involvement, so there is no category in Fig 1B for pT3 outcome because all these patients are in the lower curve for pTxN+ patients. As anticipated prospectively, we did find an inferior prognosis for patients with non-TCC histology. In all, there were 86 patients with TCC as the only histologic pattern, 39 with TCC admixed with other histologies, and 15 with only variant histology (ie, no TCC component). Cause-specific survival by histology is shown in Fig 4. Cause-specific survival for the 15 patients with non-TCC histology was statistically different from the 125 with at least some component of TCC (P = .05).
Toxicity As expected, M-VAC was quite difficult for some patients, although deliverability was better in our series than in some reports of perioperative M-VAC.6,7 Overall, 91 patients (65%) were administered at least four cycles of M-VAC. Receiving both chemotherapy and surgery was sometimes difficult. Nine patients assigned to immediate surgery never got any chemotherapy. Six either died from postoperative complications or were too ill to have chemotherapy, and three patients refused postoperatively therapy that was offered. Conversely, seven patients assigned initial chemotherapy did not come to surgery, including three patients with toxic deaths in the first two cycles of M-VAC, two patients who refused surgery, one who had rapid disease progression despite M-VAC, and one patient with financial difficulties. The time to the start of cycle 3, which should have been the same in both arms, was planned to be 84 days according to the expectations of the protocol schedule. In fact, there was a median of 103 days in the immediate surgery group and 114 days in the initial chemotherapy group. As previously reported for M-VAC, virtually all patients had some dose omissions, delays, or reductions. Even without the burden of surgery, more than 50% of patients do not receive either the day 15 or day 22 doses in at least one cycle.12 Comorbidity was significant in this cohort. Deaths because of perioperative myocardial infarction or complications of lung disease or the inability to complete planned therapy were especially prominent in older patients. In patients 70 or more years of age at registration, nine (20%) of 46 either died within the first 6 months or suffered sufficient morbidity that they could not complete at least two cycles of chemotherapy and cystectomy in any order. Such was the case for only seven (7%) of 94 patients younger than 70. Altogether, 12 patients (9%), six in each arm, experienced deaths felt to reflect toxicity of therapy. Among these 12 deaths were six related to complications of chemotherapy, two related to postoperative myocardial infarction, and four of miscellaneous postoperative complications.
Surgical Considerations
Although many have shown that it is feasible to combine chemotherapy and surgery without excessive perioperative morbidity, delivering a multiagent chemotherapy regimen such as M-VAC in the postoperative setting can be difficult. In the present trial, 54 (77%) of 70 received at least two cycles of chemotherapy after surgery. By contrast, 68 (97%) of 70 received at least two cycles among patients assigned initial chemotherapy. Interestingly, this difference was not reflected in any difference in the eventual clinical outcome of the two arms.
Relapse and Second Malignancies
Second malignancies, often also smoking-related, were fairly common in this cohort. Nine patients have had a total of 10 second primary cancers: two squamous carcinomas of the tongue, three lung cancers (two small-cell, one non–small-cell), two colon cancers, one breast cancer, one leukemia, and one glioblastoma multiforme. In all, eight patients (6%) succumbed to a second malignancy with no evidence of recurrent or metastatic urothelial cancer.
The cohort reported here was designed to be at the limit of resectability. Forty percent of these patients required a nephrostomy for obstructive uropathy at registration. More than 35% were pathologically node-positive, and 33% were 70 years of age or older. Virtually all of these patients had mutations of either p53 or retinoblastoma genes. The characterization of this cohort with respect to these and several other markers will be the subject of a subsequent report. We went to great lengths to treat patients with compromised renal function and took full advantage of the support resources of the University of Texas M.D. Anderson Cancer Center. Essentially all patients were hospitalized to receive cisplatin and forced mannitol diuresis. Stomatitis was treated aggressively, and precautions were taken to prevent methotrexate recirculation from urinary diversions. In addition, our results must also be viewed in light of the extraordinary motivation and self-selection of our patients. Although this study did not show any differences in outcome on the basis of therapy sequence, many clinical features of locally advanced urothelial cancer have been clarified by this experience. First, it is ever more apparent that the benefit of adjuvant chemotherapy demonstrated in the three randomized trials6-8 confined to a high-risk population is real. Available historical data on expected outcome from surgical series for a cohort such as the one presented here suggest optimistically that no more than 30% would be cured by means of surgery alone, and yet, by intent-to-treat, 58% of the patients we studied remain disease-free, with median follow-up that is now significantly longer than the most delayed recurrence. Of particular relevance is the finding of a nearly 40% cure rate among patients with pathologically proven lymph node metastasis (Fig 1A), better than any reported outcome with surgery alone in a similar cohort.16,17 In our view, all available evidence indicates that, in properly selected high-risk patients with clinically extravesicular bladder cancer, the addition of multiagent chemotherapy improves both cause-specific survival and cure fraction. The high rate of clinical understaging with respect to nodal involvement was a significant finding in our study and has been previously emphasized by other investigators.4,18 Clinical understaging in the context of lymphovascular invasion on the TURBx specimen was also confirmed in this study. Despite the lack of prognostic significance of lymphovascular invasion after pathologic stage is accounted for,3,5 this finding in the context of clinical staging is extremely significant and identifies a group with the same prognosis as patients with clinically extravesical disease. Clearly, the presence or absence of lymphovascular invasion on the transurethral resection material should be documented in the context of organ-preservation strategies. In contrast to the frequency of understaging, there was relatively little overstaging. Of the 44 patients judged clinically to have disease beyond the confines of the bladder and who went to surgery initially, 38 (86%) did indeed have pathologic stage T3 or higher. Thus, our results suggest that we can identify, on purely clinical grounds, the appropriate high-risk population that will benefit from available chemotherapy.
A striking finding in our study was the close correlation of down-staging to no residual muscle-invasive disease (ie, An obvious extension of combined-modality therapy, namely to those patients with even more advanced, initially unresectable cancers, has already been reported by the group at Memorial Sloan-Kettering.19 Unfortunately, patients with either pelvic sidewall fixation or gross, clinically apparent adenopathy continue to have a low cure fraction, even with the best available chemotherapy and surgery. These patients, and those who experience relapse after perioperative chemotherapy, have little hope of long-term survival with currently available treatment and speak to the need for an advance beyond the prevailing cytotoxic paradigm. These subsets represent the frontier of the clinical challenge of nonlocalized urothelial cancer.
1. Gschwend JE, Fair WR, Vieweg J: Radical cystectomy for invasive bladder cancer: Contemporary results and remaining controversies. Eur Urol 38: 121-130, 2000 2. Cheng L, Weaver AL, Leibovich BC, et al: Predicting the survival of bladder carcinoma patients treated with radical cystectomy. Cancer 88: 2326-2332, 2000[Medline] 3. Bassi P, Ferrante GD, Piazza N, et al: Prognostic factors of outcome after radical cystectomy for bladder cancer: A retrospective study of a homogeneous patient cohort. J Urol 161: 1494-1497, 1999[Medline] 4. Wijkstrom H, Norming U, Lagerkvist M, et al: Evaluation of clinical staging before cystectomy in transitional cell bladder carcinoma: A long-term follow-up of 276 consecutive patients. Br J Urol 81: 686-691, 1998[Medline]
5.
Logothetis CJ, Johnson DE, Chong C, et al: Adjuvant cyclophosphamide, doxorubicin, and cisplatin chemotherapy for bladder cancer: An update. J Clin Oncol 6: 1590-1596, 1988 6. Skinner DG, Daniels JR, Russell CA, et al: The role of adjuvant chemotherapy following cystectomy for invasive bladder cancer: A prospective comparative trial. J Urol 145: 459-467, 1991[Medline] 7. Stöckle M, Meyenburg W, Wellek S, et al: Adjuvant polychemotherapy of non–organ-confined bladder cancer after radical cystectomy revisited: Long-term results of a controlled prospective study and further clinical experience. J Urol 153: 47-52, 1995[Medline] 8. Freiha F, Reese J, Torti FM: A randomized trial of radical cystectomy versus radical cystectomy plus cisplatin, vinblastine and methotrexate chemotherapy for muscle invasive bladder cancer. J Urol 155: 495-500, 1996[Medline] 9. International Collaboration of Trialists: Neoadjuvant cisplatin, methotrexate, and vinblastine chemotherapy for muscle-invasive bladder cancer—A randomized controlled trial. Lancet 354: 533-540, 1999[Medline] 10. Logothetis C, Swanson D, Amato R, et al: Optimal delivery of perioperative chemotherapy: Preliminary results of a randomized, prospective trial of preoperative and postoperative chemotherapy for invasive bladder carcinoma. J Urol 155: 1241-1245, 1996[Medline] 11. Hall MC, Swanson DA, Dinney CP: Complications of radical cystectomy: Impact of the timing of perioperative chemotherapy. Urology 47: 826-830, 1996[Medline] 12. Sternberg CN, Yagoda A, Scher HI, et al: M-VAC (methotrexate, vinblastine, doxorubicin and cisplatin) for advanced transitional cell carcinoma of the urothelium. J Urol 139: 461-469, 1988[Medline] 13. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481, 1958 14. Frazier HA, Robertson JE, Dodge RK, et al: The value of pathologic factors in predicting cancer-specific survival among patients treated with radical cystectomy for transitional cell carcinoma of the bladder and prostate. Cancer 71: 3993-4001, 1993[Medline]
15.
Cheng L, Neumann RM, Weaver AL, et al: Grading and staging of bladder carcinoma in transurethral resection specimens. Am J Clin Pathol 113: 275-279, 2000 16. Vieweg J, Gschwend JE, Herr HW, et al: Pelvic lymph node dissection can be curative in patients with node positive bladder cancers. J Urol 161: 449-454, 1999[Medline] 17. Herr HW, Donat SM: Outcome of patients with grossly node positive bladder cancer after pelvic lymph node dissection and radical cystectomy. J Urol 165: 62-64, 2001[Medline] 18. Frazier HA, Robertson JE, Dodge RK, et al: The value of pathologic factors in predicting cancer-specific survival among patients treated with radical cystectomy for transitional cell carcinoma of the bladder and prostate. Cancer 71: 3993-4001, 1993
19.
Dodd PM, McCaffrey JA, Herr H, et al: Outcome of postchemotherapy surgery after treatment with methotrexate, vinblastine, doxorubicin, and cisplatin in patients with unresectable or metastatic transitional cell carcinoma. J Clin Oncol 17: 2546-2552, 1999 Submitted March 5, 2001; accepted June 13, 2001.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
pT2 versus pTxN+, P < .0001.
