European Urology

Introduction

Bladder cancer (BCa) is the fourth most common malignancy among men in the Western world. Almost 80% of cases are non–muscle-invasive BCa (NMIBC). Of these, tumours invading the lamina propria (T1) constitute about 20%.

The management of these tumours is under debate, but current guidelines recommend an initial bladder-sparing therapy with intravesical instillations. Bacillus Calmette-Guérin (BCG) instillation is generally regarded as the most effective treatment for patients with high-grade T1 tumours and carcinoma in situ (CIS). One of the problems with this therapy, however, is the side-effects, making maintenance therapy difficult, even impossible, in a proportion of patients. Thus, alternative schedules and drugs have been tested. One of the later possibilities could be combination therapy in the form of epirubicin plus interferon-α2b (IFN-α2b). Both drugs are effective endovesically and have few side-effects, and this combination has been evaluated already in an earlier Finnish study with good result [1]. The aim of the present prospective, randomised study on newly diagnosed T1 tumours was to evaluate whether immunochemotherapy in the above combination could be comparable in efficacy to BCG with fewer side-effects.

Patients and methods

Study design

From February 1999 to December 2006, a total of 256 patients were enrolled at 20 urologic units in Sweden (n = 206), Norway (n = 26), and Finland (n = 24) in a prospective, randomised, multicenter study conducted by the Nordic Urothelial Cancer Group. The study protocol was designed to meet the criteria of the Helsinki Declaration, including written informed consent signed by the patients. Ethics approval was granted by the Medical Faculty Ethical Committee of Umeå University (Dnr 98-145).

The inclusion criteria were patients with newly detected T1 G2–G3 urinary BCa. Good performance status (PS) was one of the conditions, as the protocol recommended cystectomy if T1 disease persisted or recurred and if progression was observed at the 6-mo follow-up. Exclusion criteria were (1) recurrent bladder tumour of any stage; (2) muscle-invasive bladder tumour at a second-look resection; (3) involvement of the urethra, prostate (ducts or stroma), or upper urinary tract; (4) hydronephrosis; (5) anticoagulation with warfarin; (6) a history of radiotherapy or systemic chemotherapy; (7) previous endovesical treatment with the investigational drugs other than a single instillation of chemotherapy, including epirubicin after transurethral resection of the bladder (TURB); (8) history of tuberculosis; and (9) immune deficiency and other malignancy (except basal cell carcinoma of the skin).

Pretreatment investigation studies included physical examination, blood analysis, cytology, bladder volume, urine culture (if necessary), and chest x-ray. An intravenous pyelogram (IVP) within the last 6 wk was also required. All patients underwent an initial transurethral resection (TUR) of all visible tumours followed by a second-look resection, including bladder mapping and biopsy of the prostatic urethra within 4–8 wk. Randomisation was by computer and through a telephone service at the Oncology Centre at Umeå University after the second look. Stratification was based on histologic grade and associated CIS.

Patients received treatment with either one ampoule (2 ml in 100 ml saline) of BCG (OncoTICE, Organon Teknika, Boxtel, the Netherlands) or the combination of 50 mg of the dry substance epirubicin (Farmorubicin, Pharmacia GmbH, Erlangen, Germany) and 10 million units (dissolved in 100 ml of saline) of IFN-α2b (Intron A, Schering-Plough, Kenilworth, NJ, USA). Both regimens were given as induction treatment for 6 wk followed by maintenance therapy for 2 yr (treatment schedule in Fig. 1).

Create Platinum Slide Series

Fig. 1 Treatment and assessment schedule.
TURB = Transurethral resection of the bladder*; R = randomisation; Pex = multiple biopsies; BCG = bacillus Calmette-Guérin; IFN = interferon-α2b; Epi = epirubicin.
* If no tumour, TURB at primary location; second look refers to TURB at the primary tumour location plus mapping.

Results

Six of the 256 randomised patients were excluded because of violation of inclusion criteria (Fig. 2). Of the remaining eligible 250 patients, 198 were men and 52 women. One hundred twenty-six patients were randomly assigned to the BCG arm and 124 patients were assigned to the combined, experimental arm. The tumour characteristics (grade, size, multiplicity, and concomitant CIS) of the two groups were well balanced (Table 1).

Create Platinum Slide Series

Fig. 2 Flow of patients.
BCG = bacillus Calmette-Guérin; Epi = epirubicin; IFN = interferon-α2b.

Regarding the primary end point, at the 6-mo follow-up, 77 patients (62%) were disease free in the combination arm as opposed to 92 patients (73%) in the BCG arm (p = 0.065). Details of the disease status at that time are shown in Table 2. At the time of this analysis, 149 patients had completed the 24-mo treatment schedule, and the observation time was a median of 15 mo (range: 1–57). The statistical analyses of DFS and progression-free survival are depicted in Fig. 3, and Fig. 4. The former showed a significant difference in favour of BCG (p = 0.012), while there was no difference with respect to progression. DFS was also analysed according to the stratification criteria (Fig. 5). The superiority of BCG over the combination therapy was significant only in the group of patients with associated CIS.

Create Platinum Slide Series

Fig. 3 Disease-free survival of patients according to randomisation arm (p = 0.01).
DFS = disease-free survival; BCG = bacillus Calmette-Guérin.

Create Platinum Slide Series

Fig. 4 Progression-free survival of patients according to randomisation arm.
DFS = disease-free survival; BCG = bacillus Calmette-Guérin.

Create Platinum Slide Series

Fig. 5 Disease-free survival in subgroups according to (a) Tis (p < 0.001), (b) grade 2 (p = 0.09), and (c) grade 3 (p = 0.055) status.
DFS = disease-free survival; BCG = bacillus Calmette-Guérin; CIS = carcinoma in situ; Epi = epirubicin; IFN = interferon-α2b.

Crossover was instituted in 10 of the BCG patients and 33 of the combination-treated patients. The corresponding numbers for other second-line therapy were 18 and 16, respectively (Table 3). A Cox proportional hazards model analysis of association with recurrence and progression status at 24 mo was performed. Significant variables for inferior outcome were treatment with the combination arm, tumour size >3 cm, initial multiple tumours, persistent T1 and/or Tis or grade 3 tumour at second look; initial concomitant Tis, age, and gender had no impact (Table 4). A corresponding analysis was performed separately in the two treatment arms (Table 5). Size was the only significant variable for BCG-treated patients, while multiplicity, status at second resection, and grade were significant in patients treated with the combination.

Adverse events

Two hundred fifty patients received at least six instillations as induction treatment and were thus included in the toxicity analysis. No significant differences were found for adverse events (nine in each group). Therapy was discontinued as a result of side-effects in 11 patients in the BCG arm after 2, 2, 2, 2, 5, 6, 6, 13, 14, 17, and 20 mo, respectively, and for two patients in the combination arm after 10 mo and 23 mo, respectively (p = 0.06). The remaining 237 patients received the maintenance therapy or were stopped because of recurrence. The frequency of patients with reported urinary problems in the questionnaire during treatment months is depicted in Fig. 6.

Create Platinum Slide Series

Fig. 6 Frequency of patients (%) with subjective urinary problems during treatment months.
BCG = bacillus Calmette-Guérin; Epi = epirubicin; IFN = interferon-α2b.

Discussion

The present study has several interesting points compared to earlier randomised, prospective trials of intravesical therapy. It includes only primary T1 tumours and it is, to our knowledge, the largest trial of this category. The management includes an obligatory second resection, an important procedure that will yield a 33–53% persistent tumour rate according to the literature [2]. The corresponding rate in this trial was 39%. The second resection procedure has also proven efficacious in reducing later recurrence rates [3].

That the T1G3 category is high risk is not controversial, as also seen in the Swedish BCa registry, with a 5-yr relative survival rate of 69% [4]. The corresponding result for patients with T1G2 was 82%, showing that this tumour category is also life threatening. Patients registered as T1G1 in that registry constituted 5% of all T1 tumours, and a review showed that all were wrongly classified. In our setting, all T1 tumours are managed similarly, and that was the reason to include the T1G2 group in the trial, as well, even if these patients sometimes are considered at intermediate risk.

BCG was more effective in reducing recurrences than the combination of chemotherapy and immunotherapy. The subgroup analysis showed that the major advantage of BCG was in those cases with concomitant CIS, which is in accordance with earlier data from our group that showed that BCG is more efficacious than single-drug chemotherapy in reducing flat lesions [5]. Patients with concomitant CIS have a modest increased risk for recurrence and a huge risk increase for progression according to the European Organization for Research and Treatment of Cancer (EORTC) risk tables [6]. However, of the trials that were the base for these tables, none had second resection and BCG maintenance, and the subgroup with T1G3 and Tis consisted of only 22 patients. Thus, it seems that the main advantage of BCG with maintenance is in that subgroup, and risk tables have to be updated with this information.

The progression rate in our trial was lower than the usually reported 30% with maintenance BCG treatment for this category of patients [7]. The reasons could be the compulsory second resection, early instituting of instillation therapy, maintenance schedule, and the recommendation for cystectomy with persisting disease after 6 mo of therapy. In our earlier experience, the median time to progression was >2 yr; thus, continued follow-up is necessary to evaluate this end point [8].

Recently, a subdivision of T1 tumours in risk categories has been proposed [9]. The proposed characteristics for high-risk patients were multifocal disease, associated Tis, tumours located in the dome and anterior wall of the bladder, and residual disease T1 at second resection. Unlike the other variables, re-TUR has not been incorporated into earlier risk factor analyses. In our multivariate analysis regarding recurrence and progression status at 24 mo, size, multiplicity, status at second resection, and grade were independent factors in addition to type of treatment. When studying the BCG treatment group, which is the most relevant as it is the standard, large size was the only risk factor for a poor outcome. The choice of a 3-cm cut-off was based on convention, as many other analyses have used that size. The Club Urológico Español de Tratamiento Oncológico (CUETO) group found independent predictors for recurrence after BCG therapy to be (1) female gender, (2) history of recurrence, (3) multiplicity, and (4) presence of associated CIS. Concerning progression, the predictors were (1) age, (2) history of recurrence, (3) high grade, (4) T1 stage, and (5) recurrence at first cystoscopy [10]. The difference to our database was the inclusion of recurrent tumours; however, we also used a shorter maintenance period. The reason for finding that associated CIS was not a risk factor in our trial could be that the international risk grouping was based mainly on trials without BCG maintenance in analogy with the EORTC risk tables, as discussed above. A recurrence in the follow-up of a high-risk tumour is an ominous sign usually leading to cystectomy. Thus, progression to a higher stage is not awaited in clinical routine. The low number of cystectomies during follow-up in our dataset is surprising and has to be further analysed.

Regarding toxicity, surprisingly, no significant differences were found. It is our opinion that the side-effects of BCG have become less of an issue as the experience of this treatment has increased. The long-term efficacy of the most commonly used drugs for the high-risk category has recently been reported in an individual patient data meta-analysis [11]. It showed that almost half of the treated patients had recurrence and thus highlighted the shortcoming of single-drug therapy. Several efforts have been made to combine different intravesical therapies with the hope of increasing antitumor efficacy [12], and [13]. Unfortunately, no evidence exists that these combinations have been superior to single-drug treatment. New drugs with other mechanisms of action, especially for the high-risk category of patients, are urgently needed.

Conclusions

For prophylaxis of recurrence, BCG was more effective than the combination of epirubicin and IFN-α2b. There were no differences regarding progression and adverse events between the two treatments.