Gemcitabine plus epirubicin in patients with advanced urothelial carcinoma who are not eligible for platinum-based regimens




The objective of this study was to evaluate the efficacy and toxicity of gemcitabine plus epirubicin in previously untreated patients with advanced urothelial carcinoma who were not eligible for cisplatin-based regimens.


Patients with advanced urothelial carcinoma and at least one of the following characteristics were eligible: impaired renal function (creatinine clearance < 60 mL per minute), an Eastern Cooperative Oncology Group performance status (PS) ≥ 2, and age ≥ 75 years. The treatment included epirubicin 70 mg/m2 as an intravenous bolus on Day 1 and gemcitabine 1000 mg/m2 over 30 minutes on Days 1 and 8 of a 21-day cycle.


Thirty-eight patients entered the study, and a total of 152 cycles were administered, with a median of 4 cycles per patient (range, 1–6 cycles per patient). The following Grade 3–4 hematologic toxicities were reported (percent of cycles): neutropenia, 22.4%; anemia, 11.2%; and thrombocytopenia, 6.5%. No cardiac, renal, or hepatic toxicities were observed. Dose intensities of epirubicin and gemcitabine were 19.6 mg/m2 per week (84%) and 532.2 mg/m2 per week (80%), respectively. There were 2 complete responses (5.3%), 13 partial responses (34.2%), 11 patients with stable disease (28.9%), and 12 patients with progressive disease (31.6%), for an overall response rate of 39.5% (95% confidence interval, 25.1–55.1). The median progression free survival (PFS) and overall survival (OS) rates were 4.8 months and 8.0 months, respectively. The 1-year survival rate was 38%, and the median PFS and OS were 6.4 months and 16.4 months, respectively, in patients with PS 0–1. Thirty patients were symptomatic: Seventeen patients (56.7%) achieved a complete response, and 5 patients (16.7%) achieved a partial symptomatic response.


At the doses given in this study, gemcitabine and epirubicin had a good tolerability profile with interesting activity in patients with advanced urothelial carcinoma who were not fit for cisplatin-based regimens. Cancer 2002;95:1444–50. © 2002 American Cancer Society.

DOI 10.1002/cncr.10860

Transitional cell carcinomas of the bladder and urinary tract are the sixth and fifth most common causes of cancer deaths in North America and in Europe, respectively.1 Prior to the development of chemotherapy, the median survival of patients with metastatic disease was ≈ 4–6 months. Several single agents (cisplatin, methotrexate, anthracyclines, ifosfamide, and mitomycin C) produce response rates of approximately 20–25%, with response durations between 4 months and 6 months and with a median survival of 7–8 months.2 On the basis of Phase II and III trials,3–5 the combination of cisplatin, methotrexate, vinblastine, and doxorubicin (M-VAC) has been accepted as the standard treatment for patients with urothelial carcinomas; M-VAC has induced response rates of 39–72% (with 9–39% complete responses) and median survivals of ≈ 12 months. Randomized trials have shown that the M-VAC regimen was superior compared with single-agent cisplatin4 and compared with the combination of cisplatin, methotrexate, vinblastine, and doxorubicin5 both in terms of response rates and survival. However, very few patients are long-term survivors,6 and the toxicities are significant: Grade 3–4 nausea and emesis, 12%; Grade 3–4 neutropenia, 24–58% (with febrile neutropenia in 10–25% of patients); severe mucositis, 13–17%; renal failure, 7%; and toxic deaths, 3%.3, 4 New agents, including gemcitabine,7–10 paclitaxel,11 and docetaxel,12 have shown activity in patients with bladder and urinary tract transitional cell carcinomas. In Phase II trials, gemcitabine has been combined successfully with cisplatin,13–15 paclitaxel,16 and paclitaxel plus carboplatin,17 producing interesting response rates and median survivals. A recently published multicenter, randomized study comparing M-VAC with gemcitabine plus cisplatin (GC) documented no differences in activity, progression free survival, and overall survival, with a tolerability profile favoring GC.18 In those trials, the patients were required to have normal renal function, the majority of them had a good performance status (PS), and very few patients were age > 75 years.

Unfortunately, many patients with urothelial cell carcinoma present with advanced age, impaired renal function, and/or poor medical conditions and require less aggressive treatments. Both gemcitabine and anthracyclines have single-agent activity against urothelial carcinoma. Epirubicin is a doxorubicin analog with reduced cardiotoxicity19 that can be administered irrespective of renal function and already has been included in several regimens.20, 22 Gemcitabine (2′2′-difluorodeoxycitidine), which is a cytosine analog that has been studied extensively in the treatment of patients with a number of solid tumors,23 is well tolerated even in elderly patients. Several Phase II studies with single-agent gemcitabine at weekly doses of 1000–1200 mg/m2 in naïve9, 10 and pretreated7, 8 patients have reported response rates of 24–29% with minimal hematologic toxicity.

Because of a lack of cross resistance and possible synergism,24 several Phase I studies have evaluated different schedules of gemcitabine plus epirubicin in patients with different tumor types.25–28 A Phase II study with a weekly schedule of gemcitabine (1000 mg/m2) and epirubicin (20 mg/m2) in previously untreated patients with metastatic bladder carcinoma recently documented a response rate of 46% (including 2 complete responses [CRs]) with Grade 3 neutropenia, thrombocytopenia, and anemia in 22% of patients, 15% of patients, and 7% of patients, respectively.29

On this basis, we conducted a Phase II study to evaluate the feasibility, toxicity, and activity of gemcitabine plus epirubicin in previously untreated patients with advanced urothelial cell carcinoma who were not fit to receive platinum-containing regimens because of advanced age, impaired renal function, and/or poor PS.


Patient Eligibility

Chemotherapy-naïve patients with histologically confirmed, locally advanced or metastatic urothelial carcinoma of the bladder, ureter, or renal pelvis with creatinine clearance < 60 mL per minute, and/or age ≥75 years, and/or Eastern Cooperative Oncology Group PS ≥ 2 were eligible for study entry. Additional eligibility requirements were measurable disease, life expectancy of at least 12 weeks, adequate bone marrow reserve (i.e., leukocyte count > 3000/μL, neutrophil count > 1500, platelet count > 100,000/μL, and hemoglobin level > 10 g/dL), adequate liver function (i.e., total bilirubin level < 1.5 mg/dL and serum alanine transferase and aspartate transferase levels < twice the upper normal limits), and written informed consent. The presence of symptomatic brain metastases or hypercalcemia were considered exclusion criteria.


All patients were treated on an outpatient basis. Gemcitabine at a dose of 1000 mg/m2 was given by intravenous (IV) infusion over 30 minutes on Days 1 and 8 of a 21-day cycle; epirubicin at a dose of 70 mg/m2 was given as an IV bolus on Day 1 of each cycle. Antiemetic prophylaxis consisted of 5-hydroxytryptamine-3 receptor antagonist plus dexametasone on Day 1 and metoclopramide 20 mg IV on Day 8. Chemotherapy was administered if, on Day 1 of treatment, the absolute granulocyte count was >1500/μL and the platelet count was > 100,000/μL; if counts were not adequate, then therapy was delayed until recovery. The doses of both drugs were administered at 75% of the planned dose if any of the following toxicities occurred: febrile neutropenia with hospitalization and/or IV antibiotics, Grade 4 neutropenia lasting > 7 days, Grade 4 thrombocytopenia lasting > 3 days or bleeding, and any Grade 3 nonhematologic toxicity (except nausea/emesis and alopecia).

Dose reductions based on Day 8 blood counts were as follows: if the granulocyte count was ≥ 1000/μL and the platelet count was ≥ 75,000/μL, then gemcitabine was administered at the full dose; if the granulocyte count was 500–1000/μL or the platelet count was 50,000–75,000/μL, then the gemcitabine dose was reduced to 50%; if the granulocyte count was ≤ 500/μL or the platelet count was < 50,000/μL, then gemcitabine was omitted. If World Health Organization (WHO) Grade < 3 nonhematologic toxicity was seen, then gemcitabine was continued at the same dose; if Grade 3 (except alopecia) nonhematologic toxicity was seen, then gemcitabine was reduced to 50%; if Grade 4 nonhematologic toxicity was seen, then gemcitabine was omitted. The planned number of cycles per patient was six cycles.

Pretreatment and On-Treatment Evaluation

Before study entry, all patients underwent a complete history and physical examination, PS score, symptom evaluation, complete blood count with differential, biochemistry, urinalysis, creatinine clearance (estimated by Cockcroft formula), abdominopelvic computed tomography scan, chest X-ray, bone scan, electrocardiogram, and echocardiogram. During treatment, patients underwent weekly complete blood counts with differential, and physical examination, PS score, symptoms, complete blood tests and creatinine clearance were repeated before each treatment cycle. Electrocardiogram, echocardiogram, and tumor measurements were repeated every three cycles.

Response and Toxicity Evaluation

Responses and toxicities were evaluated according to the WHO criteria. All the patients who had received at least one cycle of therapy were considered assessable for toxicity. A CR was defined as the complete disappearance of all clinically detectable disease measured by physical examination and/or radiographic study. A partial response (PR) required a decrease > 50% in the sum of the products of the two greatest perpendicular dimensions of all measurable lesions. Progressive disease (PD) was defined as an increase ≥ 25% in the size of measurable lesions or the development of any new lesions; all other clinical situations were defined as stable disease (SD). Responses had to be maintained for at least 4 weeks. Dose reductions and omissions were recorded for Days 1 and 8 of each cycle. Delivered dose intensity was calculated by dividing the total dose received for the number of weeks under treatment. Relative dose intensity was expressed as the ratio between the delivered dose intensity and the planned dose intensity (planned dose intensities: gemcitabine, 666 mg/m2 per week; epirubicin, 23.3 mg/m2 per week).

Statistical Analysis

Patient characteristics, toxicities, response rates, and survival were determined for all treated patients (intention-to-treat analysis). All responses were assessed by an independent radiology review. The duration of response was calculated from the date of documentation of first response to the date of first evidence of PD. The time to progression was measured from the date of the first cycle to the date of disease progression. Survival was calculated from the start of treatment to death or to the last follow-up. The Kaplan–Meier method was used to construct plots of progression free and overall survival.


Patient Characteristics

From August, 1998 to June, 2001, 38 patients with advanced urothelial carcinoma were enrolled onto the study; all patients were assessable for toxicity and response. The characteristics of the patients are listed in Table 1. The median age was 71.5 years (range, 53–80 years). Thirteen patients had a PS of 0 (34.2%), 13 patients had a PS of 1 (34.2%), 9 patients had a PS of 2 (23.7%), and 3 patients had a PS of 3 (7.9%). Eight patients had undergone nephrectomy, 14 patients were age ≥ 75 years, and 30 patients had creatinine clearance levels < 60 mL/minute. The median serum creatinine concentration was 1.34 mg/dL (range, 0.60–2.62 mg/dL), and the median creatinine clearance was 53.7 mL/minute (range, 27.8–111.1 mL/minute). Nineteen patients (50%) had more than 1 inclusion criteria. Thirty patients (78.9%) were symptomatic: 29 patients had pain, and 1 patient had dyspnea. The sites of primary disease were the bladder in 31 patients and the upper urinary tract in 7 patients. Six patients (5 with bladder carcinoma and 1 patient with carcinoma of the ureter) had locally advanced disease, and 32 patients (26 patients with bladder carcinoma, 5 patients with carcinoma of the renal pelvis, and 1 patient with carcinoma of the ureter) had metastatic disease. Twenty-seven patients had undergone prior surgery: 16 patients underwent radical cystectomy, 3 patients underwent partial cystectomy, 8 patients underwent nephrectomy, and 1 patient underwent resection of the ureter; in addition, 7 patients underwent nephrostomy. Metastatic sites were the lymph nodes in 16 patients, bone in 14 patients, the lungs in 10 patients, the pelvis in 9 patients, the liver in 7 patients, the bladder in 5 patients, soft tissues in 3 patients, the rectum in 2 patients, the ureter in 2 patients, and the peritoneum in 1 patient. Seventeen patients had 1 metastatic site, 12 patients had 2 metastatic sites, and 9 patients had ≥ 3 metastatic sites.

Table 1. Patient Characteristics
  1. ECOG. Eastern Cooperative Oncology Group.

No. of patients38
Gender (M/F)32/6
Age (yrs) 
ECOG performance status 
Creatinine clearance (mL/minute) 
Site of primary disease 
 Upper tract7
Stage of disease 
 Locally advanced6
Sites of metastatic disease 
 Lymph nodes16
No. of metastatic sites 
 ≥ 39

The patients received a total of 152 cycles of chemotherapy (median, 4 cycles; range, 1–6 cycles); responding patients received a median of 6 courses (range, 3–6 courses), and patients with SD received a median of 3 courses (range, 3–6 courses). One patient received only one course, and two patients received two courses of chemotherapy due to early death. Seven patients are still on treatment.

WHO Grade 3–4 neutropenia, thrombocytopenia, and anemia occurred in 22.4% of cycles, 6.5% of cycles, and 11.2% of cycles, respectively. Two episodes of febrile neutropenia that required hospitalization were reported. No patient required platelet transfusions because of bleeding, although nine patients required blood transfusions. No WHO Grade 3–4 nonhematologic toxicities were observed (Table 2). No toxic deaths occurred. Most of the cycles were administered without dose reductions (61%) and on the scheduled infusion day (86%). The majority of dose reductions or omissions of gemcitabine occurred on Day 8. The numbers of dose reductions or omissions of gemcitabine and epirubicin are reported in Table 3. The delivered dose intensity of gemcitabine was 532.2 mg/m2 per week (planned dose, 666.7 mg/m2 per week), and the delivered dose intensity of epirubicin was 19.6 mg/m2 per week (planned dose, 23.3 mg/m2 per week). The relative dose intensities of gemcitabine and epirubicin were 0.84 and 0.80, respectively.

Table 2. Hematologic and Nonhematologic Toxicities
ToxicityWHO grade (% of cycles)
Grade 1Grade 2Grade 3Grade 4
  1. WHO: World Health Organization.

Table 3. Dose Reductions and Omissionsa
  • a

    Total number of cycles, 152 (median, 4 cycles).

Day of administration11811
Dose adjustment (no.)   

Two patients (5.3%) achieved a CR, 13 patients (34.2%) achieved a PR, 11 patients (28.9%) had SD, and 12 patients (31.6%) experienced PD. The three early deaths were considered PD. The overall response rate was 39.5% (95% confidence interval [95%CI], 25.1–55.1).

The median duration of response was 5.7 months (range, 1.6–21.4 months). Objective responses have been observed in patients with locally advanced disease (4 patients who achieved a PR; 66.7%) and with metastatic disease (2 patients who achieved a CR and 9 patients who achieved a PR; 42.3%). Sites of disease for the responders in the subgroup of patients with metastases included the lymph nodes in six patients, the lungs in five patients, the liver in two patients, bone in three patients, the pelvis in three patients, and the peritoneum in one patient (Table 4). Thirty patients were symptomatic: 17 patients (56.7%) who achieved a CR, 5 patients (16.7%) who achieved a PR, and 8 patients (26.6%) who had SD. Twenty-five patients had a PS ≥ 1 at the time of study entry, and 16 of 25 patients (64%) experienced an improvement in their PS score at the end of therapy. The median progression free survival was 4.8 months (95%CI, 1.6–7.9 months), the median survival was 8.0 months (95%CI, 3.4–12.7 months) (Fig. 1), and the 1-year survival rate was 38%. In patients with a PS of 0–1, the median progression free survival and overall survival were 6.4 months and 16.4 months, respectively (Figs. 2,3). The median follow-up is 6.4 months (range, 1.0–33.2 months); at the time of this analysis, 15 patients (39.5%) were still alive, and 5 patients remain in remission. At the time they developed disease progression, 14 patients received further treatment with chemotherapy (carboplatin plus paclitaxel, 6 patients; docetaxel, 1 patient), radiotherapy (6 patients), or surgery (radical cystectomy, 1 patient).

Table 4. Characteristics of Responding Patients
Age (yrs)CrClPSSite of primary tumorNephrectomySite of metastatic diseaseNo. of cyclesSurvival (mos)Response
  1. PS: performance status; CR: complete response; CrCl: creatinine clearance; PR: partial response.

6574.401BladderNoPelvis, peritoneum, rectum629+CR
7553.301BladderNoLymph nodes625CR
6527.802BladderNoLocally advanced68PR
7567.200BladderNoLocally advanced610PR
7260.201UreterNoLocally advanced415+PR
7728.621BladderYesLocally advanced54PR
6835.603PelvisYesLymph nodes68PR
7054.400UreterYesLymph nodes631+PR
5375.120BladderYesLung, lymph nodes610+PR
7856.901PelvisNoLung, liver, bone44PR
7846.031BladderNoLymph nodes417+PR
6337.640BladderNoPelvis, bone, lymph nodes69+PR
7242.621BladderNoLung, liver69+PR
6655.602BladderYesLung, bone, pelvis42+PR
Figure 1.

Time to progression (TTP) and overall survival (OS). C.I. 95%: 95% confidence interval.

Figure 2.

Time to progression based on performance status (PS). C.I. 95%: 95% confidence interval.

Figure 3.

Overall survival based on performance status (PS). C.I. 95%: 95% confidence interval.


Cisplatin-based regimens are the standard treatment for patients with advanced urothelial carcinoma, and, in randomized trials, M-VAC has been confirmed as the only regimen with a potential survival benefit.4, 5 However, this regimen has substantial side effects that can be life threatening: Nausea/emesis, neutropenia, neutropenic fever, mucositis, and renal insufficiency have been reported in a significant proportion of patients, with some toxic deaths reported.3, 4, 6 In a recently published trial, a reduced incidence of hematologic and nonhematologic toxicities was reported with the gemcitabine-cisplatin combination compared with M-VAC, although only patients with a good PS and normal renal function were included.18

New combination regimens have been tested in Phase II studies. Paclitaxel plus gemcitabine16 has induced a response rate of 54% and a median survival of 14.4 months with Grade 3–4 leukopenia in 20% of courses, Grade 3 thrombocytopenia in 4% of courses, and Grade 3 anemia in 12% of courses; 19% of those patients experienced febrile neutropenia, which required 11 hospitalizations, and 1 toxic death was reported. The combination of paclitaxel plus gemcitabine and carboplatin17 has shown a high response rate (32% CRs and 36% PRs), with a median survival of 14.7 months and with Grade 3–4 neutropenia in 73.4% of patients, thrombocytopenia in 42.8% of patients, anemia in 24.4% of patients, and febrile neutropenia in 1.4% of patients.

In all of those studies the patients, were required to have normal renal function, very few had a poor PS, and the median age was ≈ 65 years. Therefore, the optimal treatment of patients with advanced age, impaired renal function, or a poor PS remains to be defined. Active single agents can induce objective responses; however, response durations generally are short.2

The current study reports the feasibility, toxicities, and activity of gemcitabine combined with epirubicin in patients with advanced urothelial cell carcinoma who are not eligible to receive platinum-based chemotherapy because of compromised renal function, poor PS, and/or advanced age. Our data demonstrate that this combination can be administered nearly at the planned doses (relative dose intensities were 0.80 and 0.84 for gemcitabine and epirubicin, respectively) without granulocyte-colony stimulating factor (G-CSF) in a population of unfit patients. This combination has produced activity (odds ratio, 39.5%) in patients with locally advanced and in metastatic disease, including liver, lung, and bone metastases; it is interesting to note that symptom control (73.4%) and an improvement in PS score (64%) also were observed. The progression free survival (4.8 months) and overall survivals (8.0 months) were inferior to those reported in other trials with more toxic regimens. However, in our study, the progression free and overall survival of patients with good PS, despite age and impaired renal function, were 6.4 months and 16.4 months, respectively, and compared favorably with the survival reported in patients on cisplatin-based trials.4, 14, 18

Recently Vaughn and et al.30 reported a 50% objective response rate with carboplatin plus paclitaxel in patients with advanced age and reduced renal function; however, this combination was not devoid of significant toxicities: 15% of patients experienced Grade 3 peripheral neuropathy, Grade 3–4 neutropenia was observed in 39% of cycles, and 21% of patients experienced at least of one episode of febrile neutropenia. G-CSF was required in 26% of cycles.

The combination of paclitaxel and carboplatin was evaluated by the Southwest Oncology Group in 29 patients, and significant neurologic and hematologic toxicities were reported along with a lower response rate (21%), which was explained by the high proportion of patients with poor prognostic features. However, those patients had a median survival of 9 months, similar to the survival observed in previous studies with the same regimen.31

In another recently published study32 that included 17 patients with poor prognostic features, carboplatin plus gemcitabine produced an overall response rate of 58.8%, a median time to disease progression of 4.6 months, and a median overall survival of 10.5 months. WHO Grade 3–4 granulocytopenia and thrombocytopenia were reported in 70% of patients and 47% of patients, respectively.

In the current study, the incidence of Grade 3–4 neutropenia was 22.4%, and only 2 episodes of febrile neutropenia that required G-CSF support with hospitalization were reported. Moreover, apart from alopecia, no Grade > 2 nonhematologic toxicity was observed.

The results of the current study have demonstrated that the combination of gemcitabine and epirubicin is an effective regimen that can be administered safely to patients with urothelial carcinoma with advanced age, poor PS, and/ or impaired renal function. A more prolonged survival can be obtained if the patients have a good PS; however, symptomatic responses and improvements in PS score were observed in 73% and 64% of patients, respectively. On the basis of activity and tolerability, this regimen deserves further development, including the addition of other active agents for patients who are in good condition.