Neoadjuvant M-Vac (Methotrexate, Vinblastine, Doxorubicin and Cisplatin) Effect on the Primary Bladder Lesion

0022-534 7/88/1393-04 70$02.00/0 Vol. 139, March

THE JOURNAL OF UROLOGY

Printed in U.S.A.

Copyright © 1988 by The Williams & Wilkins Co.

NEOADJUVANT M-VAC (METHOTREXATE, VINBLASTINE, DOXORUBICIN AND CISPLATIN) EFFECT ON THE PRIMARY BLADDER LESION HOWARD I. SCHER,*·t ALAN YAGODA, HARRY W. HERR, CORAN. STERNBERG,t GEORGE BOSL, MICHAEL J. MORSE, PRAMOD C. SOGANI, ROBIN C. WATSON, D. DAVID DERSHA W, VICTOR REUTER, NANCY GELLER, PHYLLIS S. HOLLANDER, E. DARRACOTT VAUGHAN, JR., WILLET F. WHITMORE, JR. AND WILLIAM R. FAIR From the Solid Tumor Service, Department of Medicine, Urology Service, Department of Surgery, and Departments of Medical Imaging, Pathology, and Biostatistics and Epidemiology, Memorial Sloan-Kettering Cancer Center, and Departments of Medicine and Surgery (Division of Urology), New York Hospital and Cornell University Medical College, New York, New York

ABSTRACT

Of 50 patients with bladder cancer given 1 to 5 cycles of neoadjuvant methotrexate, vinblastine, doxorubicin and cisplatin in a pilot phase I and II study 63 per cent of 41 with pure transitional cell stage T2-4 lesions responded. While significant downstaging occurred by transurethral resection of the bladder in 70 per cent and by cytology in 60 per cent of the patients, the final T response rate by all noninvasive clinical staging procedures, including sonography and computerized tomography, revealed complete remission in 24 per cent and partial remission in 39 per cent. Of 30 patients who underwent pathological staging 33 per cent achieved stage PO and 17 per cent stage Tis disease or P less than T. Despite extensive re-evaluation by transurethral resection of the bladder and other noninvasive staging procedures, a clinical staging error (T versus P) of 38 per cent was observed. Of the other 9 patients 4 with mixed nontransitional cell histological findings at presentation never achieved complete remission, although 3 had resolution of all transitional cell elements and 5 (10 per cent) were inevaluable. The toxicity of the regimen was generally acceptable but 6 per cent of the patients required hospitalization for neutropenic fever. While this active regimen can clinically (T) and pathologically (P) induce downstaging in a significant number of patients with primary bladder tumors, this pilot study has raised serious questions concerning the design of future nonrandomized and randomized neoadjuvant studies. (J. Ural., 139: 470-474, 1988) The traditional approach for stages T2-4NOMO bladder tumors is radical cystectomy or radiation therapy, either alone or in combination. However, 5-year death rates generally exceed 55 to 75 per cent with more than two-thirds of the patients dying of metastatic disease. 1- 3 The high failure rate for locally advanced (T3b-4) bladder lesions relates to tumor grade, vascular invasion and/or lymph node involvement in 40 to 60 per cent of the cases. 2• 3 Furthermore, preoperative clinical evaluation of muscle infiltrating disease, which includes examination with the patient under anesthesia plus cystoscopy and biopsy (transurethral resection of the bladder), clinically understages (T less than P) bladder cancer in more than 30 to 50 per cent of the cases. 3 With recognition that most patients with high grade, high stage bladder cancer already have systemic rather than local disease, significant improvement in survival will require effective systemic therapy. Results with methotrexate, vinblastine, doxorubicin and cisplatin (M-VAC) for advanced disease4 led to its use as initial therapy in a phase I and II trial in 50 patients with stage T24NOMO lesions to evaluate efficacy in the primary bladder site and the number of cycles needed to achieve maximal response, as well as to assess accuracy of cystoscopy and over-all clinical staging (T) by pathological (P) correlation. This paper presents data on the first 50 patients with primary bladder cancer given neoadjuvant M-VAC therapy. Accepted for publication September 29, 1987. Supported by the National Institutes of Health Grant CA-05826, and the Isadore Komanoff and Solid Tumor Service Funds. * Requests for reprints: Solid Tumor Service, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, New York 10021. t Recipients of the American Cancer Society Career Development Award.

MATERIALS AND METHODS

From December 1983 to June 1986, 50 patients with primary bladder cancer received M-VAC. Entry required documentation by the Department of Pathology at Memorial Sloan-Kettering Cancer Center (MSKCC) of a muscle-infiltrating tumor by transurethral resection of the bladder without clinical evidence of nodal (NO) or metastatic (MO) disease. Initial and followup transurethral bladder resections were done at MSKCC or New York Hospital. Re-staging was performed at MSKCC or New York Hospital in 48 patients and only 2 had pathological review of submitted cystectomy specimens. Initial and followup diagnostic radiological tests were reviewed independently (R. C. W. and D. D. D.) as was all pathological material (V. R.). To assess clinically local extent of tumor before therapy all patients underwent examination while they were under anesthesia, as well as transurethral resection of the bladder with cold cup biopsies and bladder mapping to label sites of disease and of biopsies. Surgical specimens with perivesical fat, lymph nodes, and available urethra and ureters were extensively sampled pathologically in the tumor and surrounding normal-appearing tissue areas. Other criteria for objective and subjective response, toxicity, patient entry, diagnostic tests and re-evaluation procedures were identical to those reported previously.4 The M-VAC and antiemetic schedules also were identical to those described except that no dose reduction of doxorubicin was planned, since the patients had not received pelvic irradiation. Therapeutic efficacy was evaluated on the basis of sequential changes in T category as assessed by visual and physical examination with the patient under anesthesia by cystoscopy and bimanual palpation, pathological review of transurethral bladder resection specimens (T category by the

470

NEOADJUVANT CHEMOTHERAPY EFFECT ON PRIMARY BLAD~ ER LESION

t umor, nodes and metastasis staging system of t he International Union Against Cancer when obtained at cystoscopy), T noninvasive clinical diagnostic staging [computerized tomography (CT) scan and/or ultrasound], urine cytology and, las~ly, over-all final T and final P (cystectomy or at laparotomy with biopsy of the external aspect of the bladder at the ~stim~ted site of the original tumor and selected lymph node dissect10n) evaluation. The plan was to evaluate a series of patients after 1 to 5 cycles, and 39 underwent transurethral resection of the bladder after cycle 1 or 2. . Response criteria incorporated those suggested by the First International Consensus Development Conference as discussed previously.• Downstaging by 2 or more T categories was used initially. As the trial continued it became obvious that most lesions were evaluable rather than bidimensionally measurable and only downstaging to Tis or TO could be considere? significant. Any response less than complete (CR) and partial (PR) was classified as incomplete (IR) . While the categories of clinical (c) and pathological (p) complete responses, and complete response after surgical resection of residual disease (CRs) were identical to those previously described,' a clinical partial response required a greater than 50 per cent dec~ease .in the s_ize of measurable lesions by cystoscopy and nonmvasive stagmg plus downstaging by 2 or more T categorie~, or a TO ~esion by transurethral resection of the bladder with a persistent T abnormality (for example bladder wall thickening by noninvasive staging or a positive pathological urinary cytolo~ result)A pathological partial response denoted less than 3 microscopic foci of tumor or in situ disease in the surgical (cystectomy) specimen obtained at laparotomy. As the study progressed an increasing number of patients (12 of the initial 24) refused an operation despite the physician clearly stating that the only proved therapy was radical cystectomy. Patients opted to keep the bladder because of compl~te disappearance of clinical signs and symptoms couple~ wi!h reluctance to have an ileal conduit and other compromises m quality of life. Consequently, during the latter pa~t of t?is pilot study, laparotomy with selective lymph node d1ssect10n and biopsy of the serosal surface of the bladder were recommended, and they were more acceptable to patients who understood that when no tumor was found, cystectomy would not be performed. No further therapy was given to patients with a pathological complete response. Patients with a clinical co~plete respo~se and unresectable residual disease were reclassified as havmg pathological partial or pathological incomplete responses but when all residual disease could be completely resected the final response classification then became complete after surgical resection and such cases were given 2 additional M-VAC cycles. Revie~ of all available data (table 1) indicated that 1 patient had only diffuse stage Tis disease, 1 had primary adenocarcinoma of the prostate and T4NO transitional cell carcinoma, and 2 probably had tumor that was resected completely before TABLE 1. Patient characteristics No. pts. Median age (range) Sex (M:F) Median % Karnofsky Performance Status (range) Prior intravesical therapy, No.(%): Chemotherapy lmmunotherapy Both Pre-M-VAC clinical stage, No.(%): T4 T3 T2 Tis Pathology, No.(%) : Transitional cell Ca Transitional cell Ca + adenoca. Transitional cell Ca + prostatic adenoca. Transitional cell Ca + squamous cell Ca Small cell Ca

50 61 (35-78) 45:5 80 (60-90) 6 (12) 6 (12) 6 (12)

entr ance into the protoc~ l. The latter 2 patients were not considered in the evaluatia n of response. Of note, 3 patients (6 per cent) had atypical hista logical findings in the initial bladder biopsy specimen. RESULTS

T response. T response e valuation by transurethral resection of the bladder before and after M -VAC in 44 evaluable patients revealed a 68 per cent remU;sion rate with 57 per cent achieving TO and 11 per cent Tis di1:.ease (table 2). Re-staging transurethral resection of the bladd er was performed after 1, 2, 3 and 6 cycles in 22, 17, 4 and 1 J)~tient, respectively. Of the patients with stages T2, T3 and 'I'4 disease 100, 56 and 42 per cent, respectively, had downstaging to stage TO with 55 ± 14 per cent having stage TO tumo.r after a median of 2 cycles (range 1 to 6) . In addition, 8 per cent of those with stage T4 and 12 per cent with stage T3 disease achieved Tis, while 8 per cent with stage T4 and 4 per cent wit h stage T3 had Tl cancer. Post-MVAC transurethral resection of the bladder changes were inevaluable in 6 patients beca..use of prior complete resection (2), no transurethral bladder ~section before laparotomy (2) and refusal of further therapy or evaluation (2). Of the latter patients 1 subsequently unde:i-went radical cystectomy at another institution 4 months after Protocol for a stage P3NO lesion and he is considered to have a Pathologically incomplete response. A total of 17 patients de:rnonstrated significant downstaging by transurethral bladder resection after 1 (8), 2 (8) and 3 (1) cycles and, therefore, they were given 1 to 3 more cycles to achieve TO status. With 2 cycles or less stage TO disease was observed in 24 per cent and Tis in 35 per cent of the patients but after 3 or more cycles 82 per cent had TO and 18 per cent had Tis disease (McNemar's test, Z = 3.16, p = 0.001). Thus, in these 17 responding Patients further therapy seemed to increase local tumor regression converting more to TO status as assessed by transurethral resection of the bladder. Of 4 7 patients (94 per cent) evaluable for cytology 22 of 37 (59 per cent) with an initially positive cytology result became normal, while only 1 of 10 (10 per cent) with an initially negative result subsequently had a positive finding. All 3 patients with atypical histological bladder findings (table 1) had a clinically incomplete response and in those with transitional cell plus adenocarcinoma Pathological re-staging revealed persistence of only adenocarcinoma. Surprisingly, 3 patients who presented with pure transitional cell cancer in the original biopsy specimen had histological conversion to pure adenocarcinoma, round cell tumor, and mixed transitional and adenocarcinoma in 1 case each. In 45 evaluable patients the final T response assessed by transurethral resection of the bladder and all noninvasive staging was 22 ± 12 per cent clinical complete and 43 ± 14 per cent clinical partial responses, for an over-all clinical complete plus partial response rate of 64 :±: 14 per cent (table 3). Among 4 of 10 clinical complete response patients who refused an operation and, therefore, who must remain in the final analysis as T staged only, 2 had relapse with stages Tis and T3 disease at 8 and 12 months, respectively, and 2 are free of disease for 8+ and 19+ months.

TABLE 2. Post-M-VAC T stage by transurethral resection of bladder

tumors Pre -M-VAC

14 (28) 28 (56) 7 (14) 1 (2)

46 (90) 1 1 1 1

(2) (2) (2) (2)

471

Post-M-VAC Stage

Stage

No. Entered

No. Evaluated

T4

T3

Tl

T4 T3 T2 Tis Totals

14 28 7 1 50

12 25 6 ·l

5 0 0 0

0 7* 0 0

1* 0 0

Tis No. (%)

1 (8) 3 (12) 0 1 (8) 44 5 7 2 5 (11) * Includes 1 patient each with atY]lical histological findings.

TO No.(%) 5 14 6 0 25

(42) (56) (100) (57)

_j

472

SCHER AND ASSOCIATES TABLE

3. Final T response (rate)

TABLE 4.

After M-VAC

Before M-VAC T Stage

No. Entered

No. Evaluated

T4 T3 T2 Tis Totals

14 28 7 1

12 27 5 1

50

45

Clinical Complete Remission No.(%) 2 6 2 0 10

(16) (22) (40) (22)

P versus T response to M-VAC (pathological versus clinical) After M-VAC Stage

Before M-VAC

Clinical Partial Remission No.(%)

Clinical Incomplete Remission No.(%)

5* (42) 11 * (41) 3* (60) 0 19 (42)

5 (42) 10 (37) 0 1 16 (36)

* Includes 1 patient each with atypical histological findings.

Of 19 patients with a clinical partial response 14 (69 per cent) achieved stage TO disease. All 14 patients had a clinical partial response because of bladder wall thickening only, among whom 12 had negative cytology findings, and 3 of 5 had stage Tis and 2 had stage Tl disease. Of these 19 clinical partial response patients 9 (2 originally with stage T2, 4 stage T3 and 3 stage T4 cancer) refused surgery and, therefore, they remain T staged as having a clinical partial response in the final analysis. None of 6 patients with stage TO cancer had recurrence for 7+ to 29+ months, although 3 continue to have bladder wall thickening by noninvasive staging techniques and 2 have positive cytologies. Two patients whose disease was downstaged to Tis had relapse at 9 months with N3 and M+ disease, respectively. One patient with a clinical partial response, and stage T3 transitional and squamous cell carcinoma downstaged to Tl refused surgery but radical cystectomy was performed because of an enlarging mass at 9 months. A P4NO pure squamous cell carcinoma was resected completely, and he remained with a complete response postoperatively at 16 months. There were 16 patients with a clinical incomplete response: 5 of 12 (43 per cent) with stage T4, 10 of 27 (37 per cent) with stage T3 and 1 of 1 with stage Tis disease. Three patients never had P staging immediately after re-staging, 1 of whom had a '1'4 lesion after M-VAC but cystectomy was performed 8 months later because of hemorrhage and he had a PlNO lesion. He remains with a complete response at 16+ months after resection. Another patient could not be operated upon because of chronic obstructive pulmonary disease and he is alive at 12+ months after receiving irradiation. The last patient still refuses an operation for stage T3 disease and he is alive at 6+ months. P response. P staging was performed in 32 patients (2 of whom did not undergo immediate preoperative transurethral resection of the bladder) after 1 cycle in 4, 2 cycles in 16, 3 cycles in 2 and 4 cycles in 10. Surgery consisted of radical cystectomy in 22 patients, partial cystectomy in 6 and laparotomy with selected lymph node sampling and biopsy from the serosal surface of the bladder in the area of previously known disease in 4. Stage PO disease was achieved in 31 ± 15 per cent of the patients, a rate similar for M-VAC in advanced disease. 4 Stage PO was observed in 39 per cent of 23 patients with stage T3 disease compared to only 14 per cent of 7 with stage T4 cancer (table 4). To date 1 of 10 patients with stage PO disease had relapse with a median followup of 14 months (range 7 to 24+ months). When we compared only the post-M-V AC transurethral resection of the bladder T staging immediately before P staging or T - P, 72 per cent of the evaluable cases were staged correctly (T = P), while 28 per cent were understaged clinically (T < P) and none was overstaged (T > P). When only the transurethral bladder resection findings just before radical cystectomy were considered 8 of 12 patients (66 per cent) with stage TO cancer (thus with a clinical complete response by transurethral resection of the bladder alone) indeed had stage PO disease. Of the remaining 4 patients with residual muscle infiltrating disease at radical surgery 3 also had a persistent abnormality by noninvasive staging procedures.

Stage T4 T3 T2 Tis Totals

No. Entered

No. Evaluated

P4

P3

Pl

Pis

14 28 7 1

7 23 1 1

5 1* 0 0

0 11* 1* 0

0 1 0 0

50

32

6

12

I

1 1 0 1

3

PO No.(%) 1 (17) 9 (39) 0 (0) 0 (0) 10 (31)

* Includes 1 patient each with atypical histological findings.

As previously stated 4 of 10 patients with a clinical complete response by all diagnostic procedures will remain so because no P staging was permitted; of the 6 (6 per cent) remaining clinical complete response patients 5 were reclassified as having a pathological complete response after 4 to 5 cycles, while 1 had a pathological partial response because of residual stage Pis disease. P staging was obtained in 10 of 19 (53 per cent) clinical partial response patients, of whom 2 initially had stage T4, 7 stage T3 and 1 stage T2 cancer. One patient with stage T4 disease had a pathological incomplete response but the other, whose disease was downstaged to Tis, had Pis tumor (pathological partial response) and both had a complete response after surgical resection. Of the 7 stage T3 cancer patients the disease was staged correctly in 2 (T = P) and were stage Pis, 3 had clinically understaged cancer (T < P, pathological complete response) and in 2 the disease was overstaged (pathological incomplete response), of whom 1 had only residual pure adenocarcinoma despite review of the original biopsy specimen that showed only transitional cell carcinoma. The patient with T2 cancer had a pathological incomplete response and unresectable adenocarcinoma. There were 4 clinical partial and 9 clinical incomplete responses (26 per cent) that could be converted to a complete response after surgical resection of residual disease. Eight patients (62 per cent) already had relapse at a median of 10 months (range 6 to 16 months) and 3 (6 per cent) died of disease. Two patients who had only pathologically proved microscopic disease had relapse with metastatic disease and both refused immediate postoperative chemotherapy. A complete response after resection was obtained in 16 patients with a clinical incomplete response, including 3 who presented with stage T4 and 6 with stage T3 disease. Two patients with a complete response after resection of residual tumor were found to have N4 cancer and 2 had atypical histological findings. In regard to the final T + P response rate in 41 patients with pure transitional cell tumors evaluable for response 10 were T staged as having a clinical complete, 16 a clinical partial and 15 a clinical incomplete response. In 27 patients whose disease was P staged there were 6 with a clinical complete (5 pathological complete and 1 pathological partial response) and 9 with a clinical partial (3 pathological complete, 4 pathological partial and 2 pathological incomplete) responses. All 12 clinical incomplete response cases had a pathological incomplete response. Thus, the final response rate was 30 per cent complete (4 clinical complete plus 8 pathological complete) and 28 per cent partial (7 clinical partial plus 5 pathological partial). Toxicity. Hematological toxicity was similar to that reported previously (table 5). 4 Skin necrosis owing to drug infiltration, an autonomic neuropathic condition and a flair of gouty symptoms were observed in 1 patient each, and 10 per cent had a transient increase (median 0.5 gm./dl., range 0.5 to 1.7 gm./dl.) in serum creatinine. Four patients received 1, 13 received 2, 9 received 3, 20 received 4, 3 received 5 and 1 received 6 cycles. While hospitalization for nadir sepsis was required for 6 of 50 patients (12 per cent) or 6 of 158 courses (3.7 per cent), all episodes resolved with antibiotics. Only 2 patients 78 and 76 years old had a protocol break, receiving 15 mg./m. 2 doxorubicin and 2 mg./m. 2 vinblastine in the initial cycle. All patients did not tolerate

i'l

NEOADJUVANT CHEMOTHERAPY EFFECT ON PRIMARY BLADDER LESION TABLE 5.

M-VAC hematological toxicity Cycle 1

Cycle 2

Cycle 3

Cycle 4

48 44 24 21 No. pts. White blood count nadir (cells/mm. 3 ): 2.9 3.6 2.9 3.6 Median 1.1-7.4 1.0-6.6 1.2-7.0 0.8-9.4 Range Platelet nadir (cells/mm.3): 266 289 303 251 Median 70-528 22-370 112-699 82-801 Range % receiving 100% of doses: 93 92 Days 1-2 96 95 Day 15 71 66 58 66 75 81 74 Day 22 85 Cycle interval (days): 34 35 36 Median 28-63 28-57 28-68 Range

doses on days 15 and 22; a slightly higher percentage of patients had therapy on day 22 than on day 15. One patient had a protocol break in some cycles with a 25 per cent dose reduction in all drugs owing to myelosuppression (white blood count nadir 1,500 cells per mm. 3 ) and a transient increase in the creatinine of 0.5 gm./dl. Over-all, 94 per cent of the patients received full M-VAC doses at the start of each cycle. Although cycles were to be repeated monthly the median interval was longer, 34 to 36 days, because of surgical re-staging. However, 60 per cent of the patients were re-cycled within 35 days and 80 per cent within 40 days. While the need for patients to recover from the operation or to gain hospital admission caused delays, in 6 instances delay was due to mucositis or myelosuppression. DISCUSSION

The importance of this pilot neoadjuvant M-VAC study is not that patients with pure transitional cell stage T2-4 bladder lesions can be downstaged by transurethral resection of the bladder in 70 per cent (27 of 40), cytologically in 60 per cent (21 of 35), clinically (final T evaluation) in 24 per cent (10 of 41) and pathologically (final P evaluation) to PO in 31 per cent (10 of 32) but, rather, to highlight the semeiotics of the term response and particularly the term complete remission. The ultimate question is whether large prospective randomized neoadjuvant chemotherapy trials that concentrate on the impact of such therapy on 5-year survival rates will negate the silent statistician-bladder tumor heterogeneity, patient selection and investigator evaluation. To determine the effect of chemotherapy on the primary tumor, clearly defined, strict criteria ofresponse are crucial and reliance on findings at transurethral resection of the bladder only, which frequently understage bladder tumors, is plagued by inadequate sampling. Sequential T evaluation revealed that 57 per cent (25 of 44) and 11 per cent (5 of 44) of the patients had conversion to stages TO and Tis disease, respectively, and 59 per cent (22 of 37) had conversion from positive to negative cytology findings. However, such changes may not be indicative of response, particularly when the superficial component of a tumor responds or has been excised and the deeper component remains intact below the lamina propria, or when extravesical disease, possibly unrecognized, is present. When stringent response criteria are applied, the transurethral resection complete plus partial response rate of 58 per cent decreases to a final preoperative T clinical complete response rate of only 22 per cent (10 of 45) when based on all staging procedures versus a pathological complete response rate of 31 per cent (10 of 32). Three patients had nontransitional cell components in postM-VAC surgical specimens with absence of such components in the initial pre-M-VAC biopsies. Such findings may represent post-therapy sampling error or tumor conversion. Logothetis and associates reported similar data and lack of response with the cyclophosphamide, doxorubicin and cisplatin (CISCA) regimen in patients with atypical histological findings. 5 A major problem of neoadjuvant studies is the clinical staging

473

system. In our study 8 of 30 cases (26 per cent) were still understaged incorrectly at preoperative cystoscopic evaluation despite use of sonography and CT scans. A similar problem has been demonstrated in many pre-cystectomy series that have described understaging (and to a lesser extent overstaging) in 35 to 50 per cent of the cases, particularly for stages B and C, and an error of 17 per cent in determining depth of muscle invasion in initial biopsy specimens. 6 M-VAC may be more efficacious against small volume disease with 100 per cent (6 of 6) of stage T2 versus 68 per cent (17 of 25) of stage T3 versus 50 per cent (6 of 12) of stage T4 cases achieving stage TO/Tis. Eight of 12 patients (66 per cent) whose disease was staged clinically as TO immediately preoperatively had in fact stage PO. The remaining 4 patients with a clinical partial response were identified correctly by evidence of residual abnormalities, such as bladder wall thickening on noninvasive staging procedures. In contrast, 3 of 6 patients (50 per cent) with stage TO disease and bladder wall thickening by noninvasive procedures had only fibrosis in the pathological specimen. Thus, the varied degree of resection of the primary lesion and the 35 to 50 per cent acknowledged error in the currently used clinical staging system may totally obfuscate response and survival data. For example, one study described no residual disease after transurethral bladder resection alone in 23 per cent of the cystectomy specimens. 1 An even higher incidence would be expected in patients presenting with stage T2 lesions that are small, may have a papillary component, and have been reported to be more radiosensitive and possibly more responsive to chemotherapy. 7 • 8 In 1 study that evaluated 2 cycles of cisplatin after cystoscopic resection of stages T2 and T3 lesions followed by either an operation or irradiation the reported overall survival rate was markedly influenced by the better results achieved in the T2 group. The effect on survival of complete cystoscopic resection of tumor before neoadjuvant therapies is a patient selection factor not appreciated in most studies. Several investigators also observed a favorable correlation between complete local control of stage T3 tumors after an operation or radiation therapy and patient survival. In a series of 220 patients with clinical stage T3 disease the actuarial 5year survival rate was 72 per cent in 94 (43 per cent) with a complete response to radiation therapy versus 17 per cent for partial responders and nonresponders. 9 In another radiation therapy series patients with complete disappearance of stages T2-3 lesions had a 79 per cent probability of survival versus 11 per cent for those who suffered local recurrence (p = 0.001). 10 Aggressive surgical resection, meaning full thickness of the bladder wall with adjacent perivesical fat and connective tissue, of stage T3a-3b masses also can affect response rates. In a multi-institutional study by Hall and associates such complete endoscopic (T) resection before high dose methotrexate admin istration was accomplished in 59 of 63 patients, of whom 59 per cent became clinically free of tumor and 12 per cent showed no evidence of residual invasive cancer. 11 The 3-year survival rate was similar to that obtained historically by radical cystectomy or curative radiotherapy. While patients still died of metastatic disease, excellent local control with bladder preservation was accomplished. Such results have not been reported for larger T4 tumors, since most patients die of metastases. If over-all tumor burden is important and chemotherapy is assumed to be more effective against low volume disease (that is micrometastases), then such studies could be interpreted to show an improved end result secondary only to extensive debulking either by surgery or irradiation. An alternative consideration is that so-called successful extensive debulking may simply have selected a group of patients with a more favorable pre-determined prognosis, a group without a metastatic phenotype. Is M-VAC too toxic as a neoadjuvant regimen for stages B and C cancer patients if one considers that most survive for 2 years and 40 per cent for 5 years? M-VAC was tolerated well,

474

SCHER AND ASSOCIATES

helped partly by the high expectations of physicians and patients, and the high Karnofsky Performance Status. Of the 50 patients treated 71 and 82 per cent received full dose chemotherapy on days 15 and 22, respectively, with a median cycle length of 34 to 36 days. In most cases delays were secondary to surgical re-staging. There was no evidence of a cumulative effect on hematopoiesis, hair or weight loss, and auditory and renal function. Although 12 per cent of the patients experienced nadir sepsis requiring hospitalization, none died of chemotherapy-related complications. The number of M-VAC cycles needed to achieve maximum response (PO/Pis) and whether such :response will improve survival remain to be determined. Although data are limited, in 17 patients who responded after 2 cycles and who, thus, received an additional 1 to 3 cycles, further pathological downstaging (P < T) was observed in 82 versus 24 per cent, respectively (p = 0.001). While the higher response rate could be attributed to more M-VAC therapy, the impact of selecting these responding patients to continue treatment because of significant tumor regression and of a second cystoscopic resection 2 months later cannot be dismissed. Four patients did achieve stage TO within 2 cycles and the 6 who had residual Tis disease may have had these superficial lesions resected completely during the second transurethral examination. In the remaining 7 responders with stage Tl or greater tumors resection during the second cystoscopy could possibly account for subsequent achievement of TO status. Neoadjuvant M-VAC also has been used by other investigators. Simone and Srougi evaluated 25 cases and reported a complete response in 28 per cent documented by CT scan, sonography and cystoscopy with multiple biopsies, and a partial response in 28 per cent. 12 Bukowski and associates evaluated neoadjuvant M-VAC but with cisplatin administered intraarterially in escalating doses, and reported 2 clinical complete and 2 clinical partial responses in 10 treated patients, 13 and Sen and associates also described significant clinical and pathological downstaging in all 12 patients given 2 neoadjuvant MVAC cycles. 14 Vogelzang and associates modified the M-VAC schedule, excluding the day 21 dose, and noted 3 clinical complete responders, 2 of whom had microscopic foci remaining at cystectomy (pathological partial response), 2 with Tis and 1 with progression after 2 cycles of neoadjuvant M-VAC. 15 M-VAC induces significant tumor regression of intravesical, locoregional and metastatic transitional cell carcinoma but it may not be as efficacious for existing Tis, preventing the development of future de novo Tis or invasive tumors elsewhere in the urothelium and nontransitional cell components. Not only was residual Pis disease observed in the bladder, ureter and urethra of cystectomy specimens despite complete disappearance of stage T2-4 bladder lesions, some patients who achieved a sustained clinical complete plus pathological partial response eventually had new Tis lesions. Of note, such Tis lesions -did respond to intravesical bacillus Calmette-Guerin. Although data are limited, no patient with nontransitional elements in this neoadjuvant and in the advanced disease MVAC study achieved a major response. In fact, in 3 patients who presented with mixed histological findings the transitional cell elements disappeared with only adenocarcinoma or squamous elements remaining. The complete and partial response rates were not dissimilar from that achieved with the M-VAC regimen for advanced (N3-4 or M+) disease. 4 Neoadjuvant MVAC therapy also may serve as an in vivo marker of chemosensitivity, since patients with an incomplete response and a complete response after surgical resection of residual tumor do not enter sustained remission when re-exposed to M-VAC treatment. Either additional cycles or a change in therapy may

be needed. Since 25 per cent (12) of the patients who were judged to have resectable tumors for which standard therapy would have required cystectomy retained the bladder, future trials can consider bladder preservation as an important endpoint. While this pilot study suggests that neoadjuvant M-VAC may be beneficial, such trials are complex and raise serious questions concerning study design that must be answered before the efficacy of neoadjuvant/adjuvant therapies can be defined clearly. REFERENCES

1. Richie, J.P., Shipley, W. U. and Yagoda, A.: Cancer of the bladder. In: Cancer: Principles and Practice of Oncology. Edited by V. T. DeVita, S. Hellman and S. A. Rosenberg. Philadelphia: J. B. Lippincott Co., chapt. 28, pp. 915-928, 1985. 2. Batata, M. A., Whitmore, W. F., Jr., Chu, F. C., Hilaris, B. S., Una!, A. and Chung, S.: Patterns of recurrence in bladder cancer treated by irradiation and/or cystectomy. Int. J. Rad. Oncol. Biol. Phys., 6: 155, 1980. 3. Whitmore, W. F., Jr.: Management of bladder cancer. Curr. Prob. Cancer, 4: 1, 1979. 4. Sternberg, C. N., Yagoda, A., Scher, H.1., Watson, R. C., Herr, H. W., Morse, M. J., Sogani, P. C., Vaughan, E. D., Jr., Bander, N., Weiselberg, L. R., Geller, N., Hollander, P. S., Lipperman, R., Fair, W.R. and Whitmore, W. F., Jr.: M-VAC (methotrexate, vinblastine, doxorubicin and cisplatin) for advanced transitional cell carcinoma of the urothelium. J. Urol., 139: 461, 1988. 5. Logothetis, C. J., Samuels, M. L., Ogden, S., Dexeus, F. H., Swanson, D., Johnson, D. E. and von Eschenbach, A.: Cyclophosphamide, doxorubicin and cisplatin chemotherapy for patients with locally advanced urothelial tumors with or without nodal metastases. J. Urol., 134: 460, 1985. 6. Smith, J. A., Jr. and Whitmore, W. F., Jr.: Regional lymph node metastasis from bladder cancer. J. Urol., 126: 591, 1981. 7. Slack, N. H. and Prout, G. R., Jr.: The heterogeneity of invasive bladder carcinoma and different responses to treatment. J. Urol., 123: 644, 1980. 8. Batata, M. A., Chu, F. C. H., Hilaris, B. S., Kim, Y. S., Lee, M. Z., Chung, S. and Whitmore, W. F.: Factors of prognostic and therapeutic significance in patients with bladder cancer. Int. J. Rad. Oncol. Biol. Phys., 7: 575, 1981. 9. Miller, L. S. and Johnson, D. E.: Megavoltage irradiation for bladder carcinoma: alone, postoperative, or preoperative? Proc. Natl. Cancer Conf., 7: 771, 1973. 10. Shipley, W. U., Coombs, L. J., Einstein, A. B., Jr., Soloway, M. S., Wajsman, Z., Prout, G. R., Jr. and the National Bladder Cancer Collaborative Group A: Cisplatin and full dose irradiation for patients with invasive bladder carcinoma: a preliminary report of tolerance and local response. J. Urol., 132: 899, 1984. 11. Hall, R. R., Newling, W. W., Ramsden, P. D., Richards, B., Robinson, M. R. G. and Smith, P. H.: Treatment of invasive bladder cancer by local resection and high dose methotrexate. Brit. J. Urol., 56: 668, 1984. 12. Simon, S. D. and Srougi, M.: Systemic M-VAC chemotherapy for primary treatment of locally invasive transitional cell carcinoma of the bladder (TCCB): a pilot study. Proc. Amer. Soc. Clin. Oncol., 5: 111, abstract 432, 1986. 13. Bukowski, R. M., Montie, J. E., Lee, M. and Ganapathi, R.: Neoadjuvant M-VAC with intra-arterial (i.a.) cis-platin in locally advanced transitional cell carcinoma of the bladder: phase I/II trial. J. Urol., part 2, 137: 156A, abstract 211, 1987. 14. Sen, S. E., Zincke, H., Keating, J.P. and Hahn, R. G.: Neoadjuvant chemotherapy (M-VAC) prior to cystectomy for high stage (>T2T,N,Mo) bladder cancer: do local pathological findings suggest a potential for bladder salvage. J. Urol., part 2, 137: 156A, abstract 212, 1987. 15. Vogelzang, N. J., Chodak, G. W., Awan, A. M., Ruane, M., Straus, F. H. and Schoenberg, H. W.: Neoadjuvant chemotherapy for locally invasive transitional cell carcinoma of the bladder (TCCB). Proc. Amer. Ass. Cancer Res., 28: 204, abstract 809, 1987.