Minimally important differences for interpreting the EORTC QLQ-C30 in patients with advanced colorectal cancer treated with chemotherapy.

AIM: The European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) assesses the health-related quality of life of patients in cancer trials. There are currently no minimally important difference (MID) guidelines for the EORTC QLQ-C30 for colorectal cancer (CRC). This study aims to estimate MIDs for the EORTC QLQ-C30 scales in patients with advanced CRC treated with chemotherapy and enrolled in clinical trials. METHOD: The data were obtained from three published EORTC trials that treated CRC patients using chemotherapy. Potential anchors were selected from clinical variables based on their correlation with EORTC QLQ-C30 scales. Anchor-based MIDs for within-group change and between-group change were estimated via the mean change method and linear regression, respectively, and summarized using weighted correlation. Distribution-based MIDs were also examined. RESULTS: Anchor-based MIDs were determined for deterioration in 8 of the 14 EORTC QLQ-C30 scales and in 9 scales for improvement, and varied by scale, direction of change and anchor. MIDs for improvement (deterioration) ranged from 6 to 18 (-11 to -5) points for within-group change and 5 to 15 (-10 to -4) for between-group change. Summarized MIDs (in absolute values) per scale mostly ranged from 5 to 10 points. CONCLUSIONS: These findings have clinical relevance for the interpretation of treatment efficacy and the design of clinical trials by informing sample size requirements.

Systemic review: Radiation therapy alone in medical non-operable endometrial carcinoma.

BACKGROUND AND PURPOSE: Radiotherapy is a good option for inoperable and frail patients diagnosed with endometrial cancer. Because of the lack of large multicentre trials, a systematic review was performed in an attempt to get an overview on the feasibility and efficacy of this specific approach. MATERIALS AND METHODS: We performed a bibliographic search for articles in English or French which were published in PubMed from the start of this database in January 1969 to identify publications on radiation therapy (RT) as single treatment for localised non-operable carcinoma of the endometrium. The review was completed following the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines. RESULTS: Twenty-five reports containing 2694 patients treated with RT as single treatment were identified that fulfilled the selection criteria. Disease-specific survival (DSS) at 5 years was reported for a cohort of 1322 (49.1%) patients. The combined DSS for this group of patients was 78.5% (range: 68.4-92%; 95% confidence interval: 74.5-82.5). External beam radiation therapy (EBRT) combined with brachytherapy (BT) was used in 1278 patients (47.4%), BT alone in 1383 patients (51.3%), and EBRT alone in 33 patients (1.2%). The average occurrence of grade III or worse late toxicity was 3.7% for EBRT + BT, 2.8% for BT alone, and 1.2% for EBRT alone. CONCLUSIONS: RT is in terms of disease control and toxicity, an acceptable option for non-surgical candidate patients. Prospective multicentre randomised or observational trials are needed to validate these results.

Longitudinal quality of life data: a comparison of continuous and ordinal approaches.

PURPOSE: In cancer clinical trials, health-related quality of life (HRQoL) is a major outcome measure. It is generally assessed at specified time intervals by filling out a questionnaire with ordered response categories. Despite recent advances in the statistical methodology for handling ordinal longitudinal outcome data, most users keep treating HRQoL scales as continuous rather than ordinal variables regardless of the number of categories. The purpose of this study was to compare the results of analyzing HRQoL longitudinal data under both approaches, continuous and ordinal. METHODS: The EORTC QLQ-C30 scores of two EORTC randomized brain cancer clinical trials (26951 and 26981) were analyzed using the two approaches. In the 26951 trial, a total of 368 patients were randomly assigned to receive either radiotherapy (RT) or the same RT plus procarbazine, CCNU, and vincristine. In the 26981 trial, 573 patients were randomly allocated to RT or RT plus temozolomide. Comparison of the two treatment arms was done using methods for longitudinal quantitative and longitudinal ordinal data. Both statistical methods were adapted to account for missing data and compared in terms of statistical significance of the results (p values) but also with respect to data interpretation. RESULTS: Three scales, i.e., appetite loss, insomnia, and drowsiness, presenting four response categories ("Not at all", "A little", "Quite a bite", and "Very much") were analyzed in each trial. Both statistical methods (continuous and ordinal) showed statistically significant differences between the two treatments, not only globally but also at the same assessment time points. The magnitude of the p values, however, varied at some time points and was less pronounced in the ordinal approach. CONCLUSIONS: The analysis of the two clinical trials showed that treating the HRQoL scales by a quantitative or an ordinal method did not make much difference as far as statistical significance was concerned. The interpretation of results, however, was easier under the ordinal approach. Treatment effects may be more meaningful when expressed in terms of odds ratios than as mean values, particularly when the number categories is small.

Does change in health-related quality of life score predict survival? Analysis of EORTC 08975 lung cancer trial.

BACKGROUND: Little is known about whether changes in health-related quality of life (HRQoL) scores from baseline during treatment also predict survival, which we aim to investigate in this study. METHODS: We analysed data from 391 advanced non-small-cell lung cancer (NSCLC) patients enrolled in the EORTC 08975 study, which compared palliative chemotherapy regimens. HRQoL was assessed at baseline and after each chemotherapy cycle using the EORTC QLQ-C30 and QLQ-LC13. The prognostic significance of HRQoL scores at baseline and their changes over time was assessed with Cox regression, after adjusting for clinical and socio-demographic variables. RESULTS: After controlling for covariates, every 10-point increase in baseline pain and dysphagia was associated with 11% and 12% increased risk of death with hazard ratios (HRs) of 1.11 and 1.12, respectively. Every 10-point improvement of physical function at baseline (HR=0.93) was associated with 7% lower risk of death. Every 10-point increase in pain (HR=1.08) was associated with 8% increased risk of death at cycle 1. Every 10-point increase in social function (HR=0.91) at cycle 2 was associated with 9% lower risk of death. CONCLUSIONS: Our findings suggest that changes in HRQoL scores from baseline during treatment, as measured on subscales of the EORTC QLQ-C30 and QLQ-LC13, are significant prognostic factors for survival.

Effect of completion-time windows in the analysis of health-related quality of life outcomes in cancer patients.

BACKGROUND: We examined if cancer patients' health-related quality of life (HRQoL) scores on the European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 are affected by the specific time point, before or during treatment, at which the questionnaire is completed, and whether this could bias the overall treatment comparison analyses. PATIENTS AND METHODS: A 'completion-time window' variable was created on three closed EORTC randomised control trials in lung (non-small cell lung cancer, NSCLC) and colorectal cancer (CRC) to indicate when the QLQ-30 was completed relative to chemotherapy cycle dates, defined as 'before', 'on' and 'after'. HRQoL mean scores were calculated using a linear mixed model. RESULTS: Statistically significant differences (P<0.05) were observed on 6 and 5 scales for 'on' and 'after' comparisons in the NSCLC and two-group CRC trial, respectively. As for the three-group CRC trial, several statistical differences were observed in the 'before' to 'on' and the 'on' to 'after' comparisons. For all three trials, including the 'completion-time window' variable in the model resulted in a better fit, but no substantial changes in the treatment effects were noted. CONCLUSIONS: We showed that considering the exact timing of completion within specified windows resulted in statistical and potentially clinically significant differences, but it did not alter the conclusions of treatment comparison in these studies.

Progression-free survival in advanced ovarian cancer: a Canadian review and expert panel perspective.

Ovarian cancer is leading cause of gynecologic cancer mortality in Canada. To date, overall survival (os) has been the most-used endpoint in oncology trials because of its relevance and objectivity. However, as a result of various factors, including the pattern of sequential salvage therapies, measurement of os and collection of os data are becoming particularly challenging. Phase ii and iii trials have therefore adopted progression-free survival (pfs) as a more convenient surrogate endpoint; however, the clinical significance of pfs remains unclear. This position paper presents discussion topics and findings from a pan-Canadian meeting of experts that set out to evaluate the relevance of pfs as a valid endpoint in ovarian cancer;reach a Canadian consensus on the relevance of pfs in ovarian cancer; andtry to address how pfs translates into clinical benefit in ovarian cancer.Overall, the findings and the group consensus posit that future studies should ensure that trials are designed to evaluate pfs, os, and other clinically relevant endpoints such as disease-related symptoms or quality of life;incorporate interim futility analyses intended to stop accrual early when the experimental regimen is not active;stop trials early to declare superiority only when compelling evidence suggests that a new treatment provides benefit for a pre-specified, clinically relevant endpoint such as os or symptom relief; anddiscourage early release of secondary endpoint results when such a release might increase the frequency of crossover to the experimental intervention.

Minimal clinically meaningful differences for the EORTC QLQ-C30 and EORTC QLQ-BN20 scales in brain cancer patients.

BACKGROUND: We aimed to determine the smallest changes in health-related quality of life (HRQoL) scores in the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire core 30 and the Brain Cancer Module (QLQ-BN20), which could be considered as clinically meaningful in brain cancer patients. MATERIALS AND METHODS: World Health Organisation performance status (PS) and mini-mental state examination (MMSE) were used as clinical anchors appropriate to related subscales to determine the minimal clinically important differences (MCIDs) in HRQoL change scores (range 0-100) in the QLQ-C30 and QLQ-BN20. A threshold of 0.2 standard deviation (SD) (small effect) was used to exclude anchor-based MCID estimates considered too small to inform interpretation. RESULTS: Based on PS, our findings support the following integer estimates of the MCID for improvement and deterioration, respectively: physical (6, 9), role (14, 12), and cognitive functioning (8, 8); global health status (7, 4*), fatigue (12, 9), and motor dysfunction (4*, 5). Anchoring with MMSE, cognitive functioning MCID estimates for improvement and deterioration were (11, 2*) and for communication deficit were (9, 7). Estimates with asterisks were <0.2 SD and were excluded from our MCID range of 5-14. CONCLUSION: These estimates can help clinicians evaluate changes in HRQoL over time, assess the value of a health care intervention and can be useful in determining sample sizes in designing future clinical trials.

Short-term health-related quality of life and symptom control with docetaxel, cisplatin, 5-fluorouracil and cisplatin (TPF), 5-fluorouracil (PF) for induction in unresectable locoregionally advanced head and neck cancer patients (EORTC 24971/TAX 323).

BACKGROUND: The EORTC 24971/TAX 323, a phase III study of 358 patients with unresectable locoregionally advanced squamous cell carcinoma of the head and neck, showed an improved progression-free and overall survival (OS) with less toxicity when docetaxel (T) was added to cisplatin and 5-fluorouracil (PF) for induction and given before radiotherapy (RT). The impact of the addition of docetaxel on patients' health-related quality of life (HRQOL) and symptoms was investigated. METHODS: HRQOL was assessed at baseline, at end of cycle 2, and 4, 6, and 9 months after completion of RT using the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire C30 (QLQ-C30) and the EORTC QLQ Head and Neck Cancer-Specific Module (EORTC QLQ-H&N35). The primary HRQOL scale was global HRQOL per protocol. RESULTS: Compliance to HRQOL assessments was 97% at baseline, but dropped to 54% by 6 months. Data were analysed up to 6 months. There was a trend towards improved global HRQOL during the treatment period. At 6 months after the end of RT, global HRQOL was higher in the TPF arm than in the PF arm, but the low compliance does not allow to draw definitive conclusions. Swallowing and coughing problems decreased more in the TPF arm than in the PF arm at the end of cycle 2, but to a limited extent. CONCLUSION: Induction chemotherapy with TPF before RT not only improves survival and reduces toxicity compared with PF but also seems to improve global HRQOL in a more sustainable manner.

Understanding the problem of inadequately staging early ovarian cancer.

BACKGROUND: Early ovarian cancer patients are often incompletely staged during initial surgery.(1-3) This omission can have serious adverse consequences for the prognosis of patients as the completeness of surgical staging has been identified as an independent prognostic parameter for survival.(4,5) The reasons for the problem of inadequate staging of early ovarian cancer are largely unknown. We have analysed the data of a large randomised trial in early ovarian cancer in which detailed information of the surgical staging procedure was monitored.(5) METHODS: Data of the EORTC Adjuvant ChemoTherapy In Ovarian Neoplasm (ACTION) Trial were used in which 448 early ovarian cancer patients were randomised between postoperative chemotherapy in one arm and observation following surgery in the other. In this trial strict criteria for surgical staging were advised but optimal, complete staging was performed in only 1/3 of patients. Staging characteristics of the incompletely staged patients were analysed and factors that could explain the failure to perform a complete staging were studied. RESULTS: Sampling of para-aortic nodes was omitted in 78% of the incompletely staged patients, while 52% of these patients had no pelvic lymph node dissection. Taking blind biopsies from different peritoneal sites was not performed in more than 1/3 of the incompletely staged group. Omission of the staging steps ranged from 3% (infracolic omentectomy) to 55% (biopsy of the right hemi-diaphragm). A significant difference (p=0.04) between the fraction of completely staged patients was found when comparing institutes who entered less than 5 patients (21%) versus those who included more than 20 patients (37%) in the trial. CONCLUSIONS: Even in a randomised trial in which comprehensive surgical staging was strongly advised in the study protocol the majority of patients (66%) were incompletely staged. Factors relating to a lack of surgical skills attributed most to the number of incompletely staged patients, but insufficient knowledge of the tumour behaviour and routes of spread of ovarian cancer also contributed substantially to this problem. Multicentre trials recruiting patients from many institutes with small volume contribution to the study, run the risk of inadequate adherence to the study protocol.

Phase II study on paclitaxel in patients with recurrent, metastatic or locally advanced vulvar cancer not amenable to surgery or radiotherapy: a study of the EORTC-GCG (European Organisation for Research and Treatment of Cancer--Gynaecological Cancer Group).

BACKGROUND: No standard treatment options are available for patients with advanced, recurrent or metastatic vulvar carcinoma not amenable for locoregional treatment. PATIENTS AND METHODS: In this phase II study, patients with advanced vulvar cancer received paclitaxel (Taxol) every 3 weeks for up to 10 cycles. Primary objective was response rate. Secondary objectives were response duration and toxicity. Response evaluation was assessed by World Health Organisation criteria, toxicity according to Common Toxicity Criteria. RESULTS: Thirty-one women from 10 institutions were included, with a median age of 64 (range 47-84), of which 29 were assessable for response. On study patients received a median of four cycles (range 1-10). SAFETY: Grade 3 and 4 neutropenia was seen in eight patients (8/29 = 27.6%), which in one patient resulted in neutropenic fever and treatment-related death. Further treatment-related grade 3/4 toxicity includes fatigue in three patients (10.3%) and neuropathy in one patient (3.4%). EFFICACY: Overall response was 13.8% (n = 4; two complete responses + two partial responses). With a median follow-up of 24 months, median PFS was 2.6 months (95%confidence interval 2.04-4.21). CONCLUSION: Paclitaxel shows moderate activity for local control in advanced vulvar cancer.

Phase III randomised study to evaluate the role of adjuvant pelvic radiotherapy in the treatment of uterine sarcomas stages I and II: an European Organisation for Research and Treatment of Cancer Gynaecological Cancer Group Study (protocol 55874).

The management of uterine sarcomas continues to present many difficulties. Primary surgery is the optimal treatment but the role of post-operative radiation remains uncertain. In the mid-1980s, the European Organisation for Research and Treatment of Cancer Gynaecological Cancer Group Study proposed a trial to evaluate adjuvant radiotherapy, as previous non-randomised studies had suggested a survival advantage and improved local control when post-operative radiation was administered. The study opened in 1987 taking 13 years to accrue 224 patients. All uterine sarcoma subtypes were permitted. Patients were required to have undergone as a minimum, TAH and BSO and wahsings (166 patients) but nodal sampling was optional. There were 103 leiomyosarcomas (LMS), 91 carcinosarcomas (CS) and 28 endometrial stromal sarcomas (ESS). Patients were randomised to either observation or pelvic radiation, 51 Gy in 28 fractions over 5 weeks. Hundred and twelve were recruited to each arm. The initial analysis has shown a reduction in local relapse (14 versus 24, p=0.004) but no effect on either OS or PFS. No unexpected toxicity was seen in the radiation arm. No difference in either overall or disease-free survival was demonstrated but there is an increased local control for the CS patients receiving radiation but without any benefit for LMS. Prognostic factor analysis shows that stage, age and histological subtype were important predictors of behaviour which may explain differences between CS and LMS. CS appears to show more kinship to poorly differentiated endometrial carcinomas in behaviour. LMS did not show the same benefit from radiation. These results will help shape future management and clinical trials in uterine sarcomas.

Promoter hypermethylation of FANCF and outcome in advanced ovarian cancer.

The Fanconi gene family has a role in DNA repair and inactivation of FANCF has been proposed as a mechanism of sensitisation to platinum chemotherapy. This study sought to confirm this hypothesis in cell lines and a large series of ovarian cancer samples. Promoter methylation was assessed by methylation-sensitive polymerase chain reaction of FANCF in nine ovarian cancer cell lines and 74 ovarian cancer samples taken from patients entered on a trial of cisplatin-based chemotherapy. This study confirmed methylation-dependent silencing of FANCF in one out of nine ovarian cancer cell lines. Methylation of FANCF was demonstrated in 13.2% of 53 evaluable ovarian tumour samples. Progression-free survival gave an HR of 3.63 (95% CI: 1.54-8.54, P=0.0016) in favour of the unmethylated cases. There was no association with overall survival. This study does not support methylation-dependent silencing of FANCF as a mechanism of sensitisation to platinum-based chemotherapy in ovarian cancer.

Prognostic factor analysis of health-related quality of life data in cancer: a statistical methodological evaluation.

A significant body of research exists in oncology to identify and evaluate prognostic factors, historically focused on histology, clinical stage and laboratory parameters. Recent evidence suggests that patient self-reported health-related quality-of-life (HRQOL) data provide additional prognostic information. A review by Gotay et al. of published prognostic analyses reports on the usefulness of patient-reported outcomes (PROs), including HRQOL, in predicting survival in cancer patients in clinical trials. An impressive number of studies have found a positive relationship that supports an independent association between HRQOL and survival. However, due to the considerable diversity in, for example, patient groups, types of HRQOL measures used and analytical strategies, current evidence is far from conclusive. This paper examines the statistical research methods employed, discusses key issues for HRQOL prognostic factor-analysis parameters and proposes recommendations for future outcome research.

The prognostic value of health-related quality-of-life data in predicting survival in glioblastoma cancer patients: results from an international randomised phase III EORTC Brain Tumour and Radiation Oncology Groups, and NCIC Clinical Trials Group study.

This is one of the few studies that have explored the value of baseline symptoms and health-related quality of life (HRQOL) in predicting survival in brain cancer patients. Baseline HRQOL scores (from the EORTC QLQ-C30 and the Brain Cancer Module (BN 20)) were examined in 490 newly diagnosed glioblastoma cancer patients for the relationship with overall survival by using Cox proportional hazards regression models. Refined techniques as the bootstrap re-sampling procedure and the computation of C-indexes and R(2)-coefficients were used to try and validate the model. Classical analysis controlled for major clinical prognostic factors selected cognitive functioning (P=0.0001), global health status (P=0.0055) and social functioning (P<0.0001) as statistically significant prognostic factors of survival. However, several issues question the validity of these findings. C-indexes and R(2)-coefficients, which are measures of the predictive ability of the models, did not exhibit major improvements when adding selected or all HRQOL scores to clinical factors. While classical techniques lead to positive results, more refined analyses suggest that baseline HRQOL scores add relatively little to clinical factors to predict survival. These results may have implications for future use of HRQOL as a prognostic factor in cancer patients.

Randomised trial of drains versus no drains following radical hysterectomy and pelvic lymph node dissection: a European Organisation for Research and Treatment of Cancer-Gynaecological Cancer Group (EORTC-GCG) study in 234 patients.

Drainage, following radical hysterectomy and pelvic lymph node dissection to prevent postoperative lymphocyst formation and surgical morbidity, is controversial. To study the clinical significance of drainage, 253 patients were registered and 234 patients were randomised into two arms. In one arm (n=117) postoperative drainage was performed, in the other arm (n=117) no drains were inserted. In both arms closure of the peritoneum of the operating field was omitted. The main exclusion criteria were blood loss of more than 3000 ml during surgery or persistent oozing at the end of the operation. Clinical and ultrasound or CT-scan evaluation was done at one and 12 months postoperatively. The median follow-up amounted to 13.3 months. No difference in the incidence of postoperative lymphocyst formation or postoperative complications was found between the two study arms. The late (12 months) incidence of symptomatic lymphocysts was 3.4% (drains: 5.9%; no drains: 0.9%). The difference showed a p-value of 0.06 in Fisher's Exact test. The operating time was related to the occurrence of postoperative lymphocyst formation. It was concluded that drains can be safely omitted following radical hysterectomy and pelvic node dissection without pelvic reperitonisation in patients without excessive bleeding during or oozing at the end of surgery.

Has the quality of health-related quality of life reporting in cancer clinical trials improved over time? Towards bridging the gap with clinical decision making.

BACKGROUND: Previous work highlighted a number of methodological constraints when reporting health-related quality of life (HRQOL) outcomes from randomized controlled trials (RCTs). Given this, the objective of this study was to investigate whether the quality of such HRQOL reports has improved over time. MATERIALS AND METHODS: On the basis of a predefined set of criteria, 159 RCTs with a HRQOL end point, published between 1990 and 2004 were identified and analyzed. Each study was evaluated by a number of issues (e.g. sample size and industry sponsorship) and by the "minimum standard checklist for evaluating HRQOL outcomes in cancer clinical trials". RESULTS: The quality of HRQOL reports, as measured by the overall checklist score, was independently related to more recently published studies (P < 0.0001). This relationship was independent of industry funded, HRQOL end point (primary versus secondary), cancer disease site, size of the study and HRQOL difference between treatment arms. While only 39.3% of studies published between 1990 and 2000 (89/159 RCTs) were identified as being probably robust, thus likely to support clinical decision making, this percentage was 64.3% for studies published after 2000 (70/159 RCTs). CONCLUSION: Since we found a significant learning curve in HRQOL trial reporting since 1990, it can be expected that HRQOL data will increasingly impact on clinical decision making and treatment policies in the near future.

Determinants of patient satisfaction in oncology settings from European and Asian countries: preliminary results based on the EORTC IN-PATSAT32 questionnaire.

The aim of this study was to identify factors associated significantly with hospitalised cancer patients' satisfaction with care. Patients were recruited from four geographical/cultural groups, including five European countries and Taiwan. They rated their level of satisfaction by completing the EORTC IN-PATSAT32 questionnaire at home. Additionally, data were collected on the sociodemographic and clinical characteristics and the quality of life of the patients, as well as on institutional characteristics. Of 762 patients recruited, 647 (85%) returned a completed questionnaire. The number of nurses and doctors per bed, institution size, geo-cultural origin, ward setting, teaching/non-teaching setting, treatment toxicity, global health status, participation in clinical trials and education level were all associated significantly at the multivariate level with satisfaction with doctor and nurse interpersonal skills, information provision, availability, and/or overall satisfaction. A number of patient-, institutional- and culture-related factors are associated with the perceived quality of cancer care. Future studies, with appropriate sampling frames and stratification procedures, are needed to better understand cross-national and cross-cultural differences in cancer patient satisfaction.

Phase II study of bleomycin, vindesine, mitomycin C and cisplatin (BEMP) in recurrent or disseminated squamous cell carcinoma of the uterine cervix.

OBJECTIVE: We carried out a phase II trial with BEMP [bleomycin, vindesine (Eldisine(R)), mitomycin C and cisplatin] in patients with recurrent and/or metastatic squamous cell carcinoma of the uterine cervix with the specific aim to assess whether BEMP was of particular interest when certain disease sites were involved. PATIENTS AND METHODS: Eligible patients received four cycles of E 3 mg/m(2), day 1 + 8; P 50 mg/m(2), day 1; B 15 mg/day (continuous infusion), day 2-4 and M 8 mg/m(2), day 5 (on alternate cycles), every 3 weeks during an induction phase. Thereafter, those without progression continued with MEP every 4 weeks in a maintenance phase. MEP consisted of E 3 mg/m(2), day 1 + 8, M 6 mg/m(2) (on alternate cycles) and P 50 mg/m(2), both on day 1. All drugs were given i.v. Both response evaluation and toxicity grading were assessed according to World Health Organization criteria. RESULTS: Of the 161 eligible patients, 143 were assessable for survival, 148 for toxicity and 131 for response. Overall response rate was 45% [complete (CR) 14.5%, partial response (PR) 30.5%]. Most responsive disease sites were lung, lymph nodes and skin metastases (>60% response, CR rate >25%). Median duration of response was 7.6 months. Survival was significantly better in patients with only distant metastases: 12.9 months versus 8.6 months in those with other disease sites involved (P = 0.002). In a multivariate analysis, patients with a good performance status yielded a better prognosis (P = 0.0017), as did the patients with only metastatic disease compared with those who had pelvic disease also or solely (P = 0.045). There were two toxic deaths and 21% of patients stopped treatment because of excessive toxicity. CONCLUSIONS: Patients with a good performance status and only distant metastases seem optimal candidates to receive the BEMP regimen. This benefit should be balanced against the expected serious toxic effects.