This review, based on individual patient data, evaluated the effect of adjuvant chemotherapy in invasive bladder cancer. The authors' conclusion, that there was insufficient evidence on which to base treatment decisions, seems appropriate given the small data set on which the review was based.

To evaluate the effect of adjuvant chemotherapy in invasive bladder cancer.

MEDLINE, Cancerlit and the Cochrane CENTRAL Register were searched up to and including January 1st 2001; the search terms were not reported. The references of relevant articles and reviews were also checked. In addition, the National Cancer Institute Physicians Data Query Clinical Protocols, the UK Coordinating Committee for Cancer Research trials register and the Current Controlled Trials meta-Register of trials were searched for any unpublished or ongoing trials. Searches were revised to identify any new material that had appeared by September 2004. All trialists who participated in the project were asked to help identify additional trials.

Study designs of evaluations included in the review

The review included individual patient data (IPD) from randomised controlled trials (RCTs). Trials were closed to patient accrual.

Specific interventions included in the review

Studies that compared local definitive treatment with and without adjuvant chemotherapy were eligible for inclusion. The comparison was required to be unconfounded by additional agents or interventions. The local treatment in all of the included trials was cystectomy. All trials used cisplatin-based chemotherapy, either as a single agent (1 trial) or in combination with one or more of the following agents (5 trials): methotrexate, vinblastine, cyclophosphamide, doxorubicin and epirubicin.

Participants included in the review

Patients with biopsy proven, invasive transitional cell carcinoma of the bladder were eligible for inclusion. The participants were mostly male with a median age of 62 years; tumours were predominantly pT3, grade 3.

Outcomes assessed in the review

The primary outcome was overall survival (defined as time from randomisation until death); patients still alive were censored at the date of last follow-up. Outcomes relating to overall disease-free survival, loco-regional disease-free survival and metastases-free survival were also sought. Overall disease-free survival was defined as the time from randomisation until first recurrence or progression after randomisation or death. Loco-regional disease-free survival was defined as the time from randomisation until first loco-regional recurrence or progression or death. Metastases-free survival was defined as the time from randomisation to first metastases or death. For all outcomes, death was defined as death by any cause.

How were decisions on the relevance of primary studies made?

Two reviewers independently selected articles for inclusion; any disagreements were resolved by discussion.

The integrity of randomisation (unusual patterns in the sequencing of allocation or imbalances in baseline characteristics) and follow-up (balanced by treatment arm and up-to-date) was assessed. The authors did not state how the validity assessment was performed.

The trial investigators provided IPD for their trial. The data requested included: date of randomisation, survival, local recurrence, metastases, date of last follow-up, treatment allocation, age, gender, TNM category, grade, performance status, tumour diameter, renal function and pre-treatment haemoglobin. Information was requested for all randomised patients, including those who had been excluded from the investigators' original analyses. Four of the 6 trials were stopped early, and updated follow-up was provided for all surviving patients for two of these trials and a proportion of the surviving patients from one further trial. Any queries were resolved and final database entries were verified by the responsible trial investigator or statistician.

How were the studies combined?

For each time-to-event outcome, a pooled hazard ratio (HR) with 95% confidence interval (CI) was calculated on an intention-to-treat basis for each group of trials and all trials together; a fixed-effect model was used. Absolute differences at 3 years were also calculated using the overall HRs and the control arm event rate. Survival curves were presented as Kaplan-Meier curves.

How were differences between studies investigated?

The chi-squared test and the I-squared statistic were used to assess statistical heterogeneity across groups of pooled trials. Pooled HRs were calculated for sub-groups based on chemotherapy regimen (e.g. cisplatin alone, cisplatin-based combination chemotherapy) or patient characteristics (e.g. gender, age, tumour grade).

IPD from 6 RCTs (n=493) were included.

Survival analyses were based on 283 events and 491 patients across all 6 trials. Overall, a 25% relative reduction in risk of death was shown for chemotherapy compared with the control group: the HR was 0.75 (95% CI: 0.60, 0.96), equivalent to an absolute improvement in survival of 9% (95% CI: 1, 16). No statistical heterogeneity was found. When compared with no adjuvant therapy, cisplatin as a single agent did not improve overall survival (HR 1.02, 95% CI: 0.57, 1.84). However, cisplatin-based combination chemotherapy showed a 29% relative decrease in risk of death compared with local treatment alone (HR 0.71, 95% CI: 0.55, 0.92).

Disease-free survival analyses were based on 239 events and 383 patients across 5 trials. A 32% relative decrease in the risk of recurrence or death was shown for chemotherapy compared with the control (HR 0.68, 95% CI: 0.53, 0.89), equivalent to an absolute improvement in disease-free survival of 12% (95% CI: 4, 19) at 3 years. No statistical heterogeneity was found. Cisplatin-based combination chemotherapy showed a 38% relative decrease in risk of recurrence or death compared with controls (HR 0.62, 95% CI: 0.46, 0.83).

Locoregional disease-free survival and metastases-free survival analyses were not presented since only limited data were available.

There was no evidence of any difference in the effectiveness of chemotherapy across subgroups based on patient characteristics.

There was insufficient evidence on which to base treatment decisions; further research into the use of adjuvant chemotherapy in invasive bladder cancer, in large RCTs, is needed.

The review question was supported by well-defined inclusion and exclusion criteria. The literature search was thorough and attempts were made to identify unpublished material, thereby reducing the likelihood of publication bias. A collaborative group of trial investigators was established to maximise the retrieval of IPD. The validity of the eligible trials was assessed by checking the raw data from each trial and resolving any problems with the trial investigator. The statistical analysis undertaken appeared appropriate, and heterogeneity was investigated. The authors acknowledged that the review was limited by a small data set, and fell short of the 900 events needed to reliably detect a 9% absolute survival benefit with 80% power. This was a well-conducted review and the authors' conclusion is likely to be reliable.

Practice: The authors suggested that the use of adjuvant chemotherapy should be encouraged only in the context of ongoing trials.

Research: The authors stated that further research into the use of adjuvant chemotherapy in invasive bladder cancer, in appropriately sized RCTs, is needed.

British Medical Research Council.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.