Cystectomy is the standard treatment for patients with infiltrating bladder cancer, but conservative treatment with cystoscopic resection followed by radiochemotherapy may be an alternative for highly selected patients. The addition of gemcitabine to cisplatin and radiotherapy may enhance disease control.
The long-term clinical outcomes of 26 patients enrolled in a previously published dose-finding study and a prematurely discontinued phase 2 trial were evaluated. All the patients underwent transurethral tumor resection followed by a radical dose of external radiotherapy administered at the same time as cisplatin and weekly gemcitabine therapy.
After a median follow-up of 74 months, the projected 5-year clinical outcomes were a 70.1% overall survival rate, a 78.9% disease-specific survival rate, and a 73.8% bladder-intact survival rate.
Radical cystectomy remains the gold standard of treatment for muscle-invasive bladder cancer,1 with a 5-year local control rate of 80% to 90% and 5-year overall survival rate of 40% to 60%.2-5 However, despite advances in surgical techniques such as nerve-sparing and orthotopic bladder construction, this major surgical technique has a considerable impact on the patient's quality of life. Furthermore, many patients refuse surgery or are considered unsuitable candidates because of their age or comorbidities.
Bladder preservation has been proposed as an alternative approach in highly selected patients. The use of a trimodal strategy that combines the possibility of controlling the disease locally by means of an organ-preserving transurethral resection of the bladder (TURB) followed by external beam radiotherapy (XRT), with systemic chemotherapy theoretically eradicating any micrometastases that may already be present at the time of initial diagnosis, has led to local control and survival rates that are equivalent to those obtained with cystectomy,6-8 and has the quality-of-life advantage of leaving patients with their own functioning bladder.
Several chemotherapeutic strategies have been used in an attempt to improve the outcomes of conservative approaches. No significant benefit has been gained by administering neoadjuvant treatments before radiotherapy,9 and therefore they are no longer routinely proposed, but some latest-generation drugs have more recently been administered concomitantly with external irradiation because of their radiosensitizing activity. Gemcitabine appears to be a particularly promising radiosensitizing agent because it is also active against advanced bladder cancer10 and acts synergistically with cisplatin.
We published the results of what to our knowledge was the first (dose-finding) study of gemcitabine administered with XRT and cisplatin in the context of a bladder-conserving strategy in 200311 and, given the promising local response rate, started to test the activity of the schedule in a prospective phase 2 trial, which was prematurely closed because of the low accrual rate. In the current study, we describe the long-term outcomes of the patients enrolled in the 2 trials.
MATERIALS AND METHODS
Our single-institution dose-finding study, which was initiated in March 1999 and ended in July 2002 after the enrolment of 16 patients, allowed us to establish the maximum tolerated dose (MTD) of gemcitabine administered with XRT and cisplatin. The subsequent phase 2 trial was planned as a multicenter trial involving another 9 Italian hospitals, but these institutions enrolled only 1 patient; we enrolled 9 patients but in December 2004 the policy of the Urology Department was changed and radical cystectomy was considered the only proposable treatment for invasive bladder cancer. The trial was therefore terminated after recruiting a total of 10 patients, but herein we describe only the 9 patients enrolled by us.
The patient selection criteria were the same for both studies. After the studies had been approved by the local Ethics Committees of the participating hospitals, all patients underwent cystoscopy and macroscopically complete TURB. They all had a histological diagnosis of pT2 or higher transitional bladder carcinoma without any physical or radiographic signs of metastatic disease. Before treatment, all patients underwent a physical examination, routine laboratory tests (including complete blood cell counts and electrolyte, blood urea nitrogen, creatinine, and glucose levels), liver function studies, abdominal computed tomography (CT) scan, and a chest x-ray. The inclusion criteria were an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of at least 1, and an adequate bone marrow reserve (defined as a pretreatment white blood cell count of ≥3000/μL, hemoglobin levels of ≥11.0 g/dL, and a platelet count of ≥100,000/μL), alkaline phosphatase and alanine aminotransferase levels ≤3 times the upper limit of normal, total bilirubin ≤3.0 mg/dL, and serum creatinine ≤1.3 mg/dL. Patients who were pregnant or breastfeeding were excluded, as were those with a history of previous chemotherapy or radiotherapy, or a history of malignancy other than adequately treated basal cell or squamous cell skin cancer or in situ cervical carcinoma. The enrolled patients were required to have a life expectancy of >12 weeks, be at least aged 18 years, and provide a written informed consent form.
The study protocol involved the delivery of 54 grays (Gy) of fractionated XRT (1.8 Gy/fraction × 30 fractions) over 6 weeks in combination with 6 weekly injections of gemcitabine plus cisplatin at a dose of 100 mg/m2 given on Days 1 and 22. There were 4 gemcitabine dose levels (from 200 to 500 mg & sold;m2/week, escalating by 100 mg/m2/week from 1 level to the next), and a minimum of 3 patients had to be treated at each dose level in the absence of dose-limiting toxicities. If any dose-limiting toxicity occurred in 1 of these 3 patients, a minimum of 3 additional patients were to be studied at the same dose level. The MTD of gemcitabine was defined as the dose associated with febrile neutropenia, grade 4 thrombocytopenia, grade 3 or 4 enteric toxicity, or grade 4 nonhematologic toxicity in 33% of the patients treated at that dose level. XRT-related toxicities were assessed on using the Radiation Therapy Oncology Group criteria, and chemotherapy-related toxicities were assessed using the World Health Organization criteria.
Phase 2 study
On the basis of the results of the dose-finding study, gemcitabine was administered at a dose of 400 mg/m2 on Days 1, 8, 22, and 29 together with cisplatin at a dose of 100 mg/m2 on Days 1 and 22, and XRT was administered in the same way as in the previous study.11
XRT was initiated within 6 weeks of complete TURB. During the first 4 weeks of XRT, a fractionated dose of 36 Gy (1.8 Gy/fraction × 20 fractions) was administered to the small pelvis (from S1 to the obturator foramen) using a 4-field box, 3-dimensional conformal technique; the clinical target volume included the bladder and pelvic lymphatics. During the fifth and sixth weeks, the clinical target volume included only the initial tumor volume, with the use of reduced-box or opposed lateral or anteroposterior fields. The patients were treated in a supine position with an empty bladder. Treatment was planned using a CT scan made in the treatment position, with a 1.5-cm margin being added to the clinical target volume to allow for geometric uncertainties (planning target volume). XRT was delivered using 15-megaelectron volt beams of a linear accelerator. The XRT modalities were the same in the dose-finding study and the phase 2 trial.
Chemotherapy was initiated on the first day of XRT. Cisplatin was given at a dose of 100 mg/m2 on Days 1 and 22. Gemcitabine was administered intravenously at the previously described doses on Days 1, 8, 15, 22, 29, and 36 (in the dose-finding study) or on Days 1, 8, 22, and 29 (in the phase 2 trial) during the hour preceding XRT (except for Days 1 and 22, when it was given before cisplatin). No additional doses were given after Day 36 if any doses had been withheld during the planned course of treatment.
No dose modification was planned if neutrophil counts were ≥1500/μL and/or platelet counts were ≥100,000/μL. If the neutrophil counts were between 1000 and 1500/μL, the drugs were administered at full doses with 2 subcutaneous doses of granulocyte–colony-stimulating factor (5 μg/kg/day) being administered 3 days after treatment; in the case of lower neutrophil counts, this treatment was omitted. Dose reductions of 75% for gemcitabine and 50% for cisplatin were planned if platelet levels were between 75,000/μL and 99,000/μL, and a week's rest was planned if these levels were <75,000/μL. In the case of increased creatinine levels, the treatment was stopped until forced intravenous hydration had returned them to normal values.
Evaluations During Therapy
Routine laboratory tests (complete blood counts with differential, electrolytes, glucose, and renal and hepatic function) were performed weekly during therapy. The patients were clinically assessed each week by medical oncologists and radiotherapists in a multidisciplinary setting. Any decision to reduce the dose, withhold therapy, or proceed on schedule was made on the basis of the patient's clinical status and a review of the laboratory results.
Six to 8 weeks after completing the concomitant radiochemotherapy, the patients underwent a cystoscopic re-evaluation with multiple biopsies of the initial tumor site. No further treatment was planned for patients who were free of tumor, whereas those with noninvasive and well-differentiated superficial tumors were to be treated conservatively by means of TURB plus intravesical chemotherapy or immunotherapy. The patients with deeply infiltrating or poorly differentiated residual tumors underwent cystectomy, unless major surgery was contraindicated or there were progressive distant metastases. All the patients also underwent abdominal CT scans and a chest x-ray. The follow-up procedures for local disease control assessment in all of the patients who were tumor free was comprised of alternating urinary cytology and cystoscopy every 3 months for the first 3 years, and every 6 months from the fourth year onward. In the case of local (invasive or noninvasive) disease recurrence, the proposed treatment was the same as that described earlier. The patients underwent abdominal CT scans and a chest x-ray every 6 months for the first 3 years, and yearly thereafter.
Actuarial survival probabilities were estimated using the Kaplan-Meier method12 and the following definitions.
For overall survival (OS), events were defined as death because of any cause. The time to OS was the interval between treatment initiation and death, or the most recent follow-up if no event occurred.
For disease-specific survival (DSS), events were defined as death attributable to bladder cancer. The time to DSS was the interval between treatment initiation and death because of bladder cancer, or the most recent follow-up if no event occurred.
For bladder-intact survival (BIS), events were defined as cystectomy for any reason or death because of any cause. The time to BIS was the interval between treatment initiation and cystectomy or death (whichever was shorter), or the most recent follow-up if no event occurred.
Local disease-free survival
For local disease-free survival (lDFS), events were defined as any recurrence of superficial or infiltrating bladder cancer. The time to lDFS was the interval between treatment initiation and the first bladder failure, or the most recent follow-up if no event occurred.
Distant disease-free survival
For distant disease-free survival, events were defined as distant failure (ie, metastases). The time to dDFS was the interval between treatment initiation and the occurrence of distant metastases, or the most recent follow-up if no event occurred.
For event-free survival (EFS), events were defined as any failure, death because of any cause, or cystectomy. The time to EFS was the interval between treatment initiation and the first failure, death, or cystectomy (whichever was shortest), or the most recent follow-up if no event occurred.
The 26 enrolled patients had a median age of 67.5 years (range, 51-80 years) and a median ECOG PS of 0 (range, 0-1). The details with regard to the number of patients enrolled at each dose level in the dose-finding study have been previously published.11
The weekly gemcitabine administration was generally well tolerated and was associated with manageable toxicity up to the dose of 500 mg/m2, at which time 1 intestinal perforation (favorably treated surgically) was observed 2 weeks after the end of treatment, as was 1 death because of untreated diarrhea appearing on treatment Day 30 and not reported by the patient. The major toxicities at the lower dose levels were hematological and usually occurred on Days 15 and 36.
These toxicities led to the schedule proposed in the phase 2 trial, with the gemcitabine MTD being defined as 400 mg/m2 and doses on Days 15 and 36 being omitted. A more detailed description of the acute toxicities observed during the dose-finding study has been published elsewhere.11
Phase 2 trial
During radiochemotherapy, the toxicities were manageable because there were only 2 cases of grade 3 neutropenia. Local toxicity was mild (grade 2 urinary and bowel toxicity). There was no case of radiotherapy being discontinued because of toxicity; chemotherapy was delayed in 2 cases because of hematological toxicity.
No relevant toxicities were observed during the follow-up period. The treatment did not produce any relevant late urinary or rectal toxicity in the patients who retained their bladder: only transient grade 1 to 2 irritative urinary symptoms were observed in the 6 months after the end of XRT. All these patients were still alive without cystectomy at the time of last follow-up, and retained their own bladder with normal organ function. The urologists did not encounter any relevant technical problems during surgery in the patients who underwent salvage radical cystectomy.
Fifteen patients in the dose-finding study (1 died during treatment) and all 9 in the phase 2 trial were evaluable for response; none experienced rapid disease progression before the completion of the treatment. All the evaluable patients were disease free at the time of the cystoscopic re-evaluation, and the macroscopic absence of visible lesions was confirmed by biopsies obtained at the site of the primary tumors (a response rate of 100%). Post-treatment CT scans did not reveal the appearance of any distant metastases (Table 1).
MTD indicates maximum tolerated dose; NR, not reported.
Median follow-up, mo
5-y overall survival, %
5-y disease-specific survival, %
5-y bladder-intact survival, %
5-y local disease-free survival, %
5-y distant disease-free survival, %
5-y event-free survival, %
After a median follow-up of 74 months (range, 13-124 months), 4 patients had developed an infiltrating bladder disease recurrence after 5 months to 13 months (median, 7.5 months). Salvage surgery was not proposed in 1 case because of a concomitant extravesical pelvic recurrence. Cystectomy was performed in 2 cases; the remaining patient refused surgery and died of causes unrelated to cancer 23 months after the completion of treatment. An additional 4 patients developed noninfiltrating tumors: 1 was managed with intravesical therapy and 2 underwent cystectomy; the fourth patient refused any further treatment or disease re-evaluation and remained asymptomatic and in good health after 113 months. Four patients developed distant metastases after a median of 13.5 months (range, 10 months-80 months).
The projected 5-year clinical outcomes are an OS rate of 70.1%, a DSS rate of 78.9%, a BIS rate of 73.8%, an lDFS rate of 65%, a dDFS rate of 87.5%, and an EFS rate of 53.6%. In the case of the 12 patients treated at the MTD, the projected 5-year clinical outcomes are an OS rate of 73.3%, a DSS rate of 82.5%, a BIS rate of 80.2%, an lDFS rate of 72.9%, a dDFS rate of 91.7%; and an EFS rate of 56%.
The long-term outcomes described in the current study confirm that the addition of gemcitabine to concomitant cisplatin and XRT leads to good disease control in patients with infiltrating bladder cancer.
Gemcitabine is a multitarget antifolate drug whose characteristics make it the drug of choice in the context of a multidisciplinary clinical strategy aimed at sparing the bladder in patients with transitional infiltrating cancer. It synergistically acts with cisplatin, something that reportedly led to good results in the treatment of non-small cell lung cancer, for which this combination is reportedly one of the most active therapies for patients with advanced disease.13 It is active against bladder cancer,10 and has radiosensitizing activity.14
Our group first reported the feasibility of adding gemcitabine to XRT after TURB in 2003, and other subsequently published studies have confirmed our findings (Table 2).11, 15, 16 However, our MTD findings were quite different from those of the other studies in that we observed manageable toxicities requiring only the omission of some drug doses (usually in the third week) in all of our patients who received the MTD or less (the dose recommended by us was 400 mg/m2 on Days 1 and 8 every 3 weeks), but the other studies recommended much lower doses, ranging from 27 mg/m2 twice a week15 to 150 mg/m2 weekly.16 This clear contrast most likely reflected some differences in terms of treatment (XRT technique) and study design (the definition of dose-limiting toxicity).
MTD indicates maximum tolerated dose; NA, not applicable; XRT, radiotherapy; Gy, grays; NR, not reported.
27 mg/m2 twice a wk
52.5 Gy (conformational)
No. of patients
Median follow-up, mo
Response rate (at cystoscopy)
Another significant difference was the therapeutic strategy adopted in the different trials. The other groups evaluated gemcitabine alone as a radiosensitizer and avoided the use of cisplatin during XRT, whereas we added gemcitabine into one of the most widely used radiochemotherapy schemes for the conservative management of patients with infiltrating bladder cancer, with cisplatin at a dose of 100 mg/m2 being given at the same time as XRT. Obviously, this raises the question of the role of cisplatin and its true usefulness in disease control.
All 3 dose-finding studies reported interesting response rates, with the post-treatment complete remission rate ranging from 87.5%16 to 100%.11 Given the dose-finding nature of the studies and their consequently small study populations, these results need to be considered cautiously. However, the availability of data concerning the long-term clinical outcomes of the patients enrolled in dose-finding trials may help. The University of Michigan group recently updated its findings after a median follow-up of 67 months,17 and comparison of time-dependent clinical outcomes (Table 1) indicated quite similar results. The only major difference appeared to be with regard to the projected 5-year BIS rate, which was reported as 62% in the study by Oh et al and 73.8% in ours. This is even more evident in the subsequent time interval, with survival in the US study falling to 0 after 8 years, but remaining at 60.6% for 10 years in ours (Fig. 1).
Once again, the small number of patients does not allow any definite conclusions to be drawn, but the difference suggests that the addition of cisplatin may have offered some advantage. Furthermore, the data from the current study cannot support the addition of both cisplatin and gemcitabine to XRT as being superior compared with cisplatin alone. These problems may be clarified only by formal comparative studies aimed at establishing whether XRT should to combined with cisplatin alone, gemcitabine alone, or both drugs.
However, the difficulties we experienced when enrolling patients for our phase 2 trial suggest that such studies may never be completed and highlight the cooperation of urologists as the main limitation. From a urologist's point of view, a bladder-sparing strategy should be restricted to patients at high surgical risk, with cystectomy being the only truly life-sparing therapy in the other patients. Nevertheless, the discouraging long-term survival data of a large series of patients treated surgically2-5 or radiochemotherapeutically6-8 clearly indicate that patients are the only losers in this battle for therapeutic supremacy. We strongly believe that cystectomy is the treatment of choice of patients with infiltrating bladder cancer, but also that conservative multimodal TURB, XRT, and CT management should be offered to highly selected patients because it has a clear physiological and psychological advantage that cannot be reduced by new bladder-removing surgical techniques.
In conclusion, our long-term follow-up data concerning a series of patients participating in 2 prospective trials confirm that the addition of gemcitabine to XRT and cisplatin after TURB is safe and provides good local and distant disease control. The relatively low XRT dose used in our multimodal experience may explain the good long-term functional results, and the concomitant administration of cisplatin may explain the good rate of long-term organ preservation. Conducting confirmatory and comparative trials could satisfy an unmet need but requires the multidisciplinary cooperation of urologists in selecting the right patients for a bladder-sparing strategy.