Roche trials database

Protocol number:

ML19309

Title of Study:

An open-label study to assess the anti-tumor activity of Avastin in combination with fotemustine as first-line therapy in patients with metastatic melanoma

Sponsor:

Hoffmann-La Roche

Company division:

Pharmaceutical

Product name:

Avastin

Generic name:

bevacizumab [Avastin]

Therapeutic area:

• Malignant Melanoma

Clinical study summary:

This multi-center study was designed to assess efficacy and safety of a combination of Avastin (bevacizumab) and fotemustine as first-line therapy in patients with metastatic melanoma. Patients received Avastin until disease progression, unacceptable toxicity, patient's or physician's decision, and fotemustine for up to 6 cycles. The primary efficacy endpoint was reached (Study Stage 1), Study Stage 2 was not performed.

Study center(s):

4 centers in Italy

Phase of development:

II

Objectives:

The primary objective was to assess the preliminary antitumor activity in terms of overall response rate (complete and partial responses) of bevacizumab in combination with fotemustine.

The secondary objectives were to evaluate the toxicity and safety profile of the drug combination, and to evaluate duration of response (DR), time to progression (TTP), time to treatment failure (TTF), and overall survival (OS).

Methodology:

This was a multicenter phase II study. Once informed consent was given by the patient, screening was completed and eligibility was verified, patients were registered by means of an eligibility screening form sent to the Sponsor. The trial followed a two stage optimal design: 18 patients were included in the first stage. The patient recruitment was temporarily stopped to allow the interim analysis of the primary efficacy parameter, as well as of safety data. The primary parameter was the overall response rate, based on RECIST criteria, combined with the safety evaluation, based on the adverse events recorded. If the efficacy and safety results of the interim analysis would have been positive, then an additional 18 patients would have been included, for a total of 36 patients.

Number of patients (planned/analyzed):

20 patients enrolled and treated

Diagnosis and main criteria for inclusion:

Adult patients with stage IV melanoma, previously untreated for metastatic disease with chemo- or immunotherapy. Histological or cytological confirmed cutaneous malignant melanoma; Advanced, inoperable nonchoroidal stage IV melanoma, including unresectable regional lymphatic disease and/or extensive in transit recurrent disease.

Test product, dose and mode of administration or test procedure:

Bevacizumab 15 mg/kg every three weeks intravenously until disease progression, unacceptable toxicity, patient's or physician's decision

Fotemustine 100 mg/m², intravenous infusion given on days 1, 8 and 15 (induction phase), to be repeated after 4 weeks every 21 days (maintenance phase), up to 4 to 6 cycles.

Duration of treatment:

Until disease progression, unacceptable toxicity, patient's or physician's decision

Reference therapy, dose and mode of administration or reference procedure:

N/A

Criteria for evaluation (efficacy, safety):

Efficacy

Primary: Proportion of responders and clinical benefit (assessment of overall tumor response according to RECIST criteria)

Secondary: Overall survival (OS), Time to progression (TTP), Time to treatment failure (TTF),  Time to response (TTR), Response duration (RD).

Safety: Adverse events (AE) and severe adverse events (SAE);laboratory parameters; vital signs; physical examination.

Statistical methods:

The main endpoint of the study was the best tumor response at any time during therapy. According to a classical two stage optimal design, with less than 10% considered as an unacceptable response rate and more than 30% acceptable, for 5% significance level and 90% power, 18 patients were needed in the first stage and if 2 or less had an objective response (and/or grade 3-4 clinical toxicity ≥40% was registered), then the trial would have to be closed because of poor response (and/or poor tolerability). If 3 or more patients responded (and grade 3-4 clinical toxicity <40% was registered), the trial would go to the second stage by recruiting other 18 patients to reach a total of 36. If 7 or more patients responded, it would be concluded that a response rate of 30% was possible and that a response rate of 10% was unlikely. Exact binomial methods were used to estimate the response rates and their 95% confidence intervals. Time to progression or death was estimated by the product-limit method of Kaplan and Meier.

Summary (efficacy, safety, other results):

Efficacy

Relatively to Intent-to-treat (ITT) population, only a patient (5%) reached confirmed complete remission during study period; 2 subjects (10%) achieved partial remission, 10 patients (50%) persisted in a state of stable disease and 6 individuals (30%) suffered from progression of disease. In the Per Protocol (PP) population, no patient achieved complete or partial remission; 6 patients out of 8 (75%) reported a best overall response of stable disease, while 2 of them (25%) had disease progression.

Proportion of responders

Proportions of objective responders were 0.15 (3 patients) (CI at 95% [0.03;0.38]) and 0 for ITT and PP populations, respectively.

Clinical benefit

Proportions of patients with clinical benefit were 0.65 (CI at 95% [0.41;0.85]) and 0.75 (CI at 95% [0.35;0.97]) for ITT and PP populations, respectively.

Overall survival (OS)

Relatively to ITT population, 12 patients out of 20 (60%) experienced, during study period, the failure event of death. Eight (40%) patients alive at the end of the study. In the PP population, the percentages of failed and censored were 50% and 50% respectively (with 4 patients in each “category”). Fifty percent of patients died before the 615th and the 761st day for ITT and PP populations, respectively. In ITT set, 796 days after start of treatment only 25% of patients was still alive. The mean time of survival was 532.49 days and 591.83 days for ITT and PP populations, respectively.

Time to progression (TTP)

Relatively to ITT population, 14 patients (73.68%) experienced during study period the failure event defined as occurrence of death or progression disease. Moreover, 5 patients (26.32%) could be considered censored, i.e. alive and with tumor assessment valid and available at the end of observed period. In PP population, the percentages of failed and censored were 75% and 25% respectively (with 6 and 2 patients within each “category”). Both in ITT and in PP population, the mean time to progression was about a year (365.26 days and 358.5 days, respectively, for the two analysis groups). Twenty-five percent of patients experienced death or progression of disease after 74 and 95 days from the starting day of response evidence in ITT and PP populations, respectively. Median times to progression were 249 and 232 days for ITT and PP populations.

Time To treatment Failure (TTF)

Both in ITT and in PP population, all patients had at least one negative occurrence among those taken into account in evaluation of TTF (death or discontinuation due to reasons other than ‘Protocol violation’). No patients were identified as censored. All patients recruited for the analysis failed in less than one year from the start treatment date both in ITT and in PP population. Mean times to failure were respectively 115.47 and 111.8 days for the two analysis subgroups. Twenty-five percent of patients experienced the earliest failure event after only 43 and 71.5 day from the start of treatment in ITT and PP population respectively. Median times to failure were 126 days for both ITT and PP subgroups. No patient was event-free after 6 month from the beginning of treatment.

Time To Response (TTR)

Relatively to ITT population, 1 patient (5.26% of patients recruited) reached CR; 3 individuals among the enrolled reached CR or PR (15.79% of patients). Patients considered censored for time to overall complete response and time to overall response were, therefore, 18 (94.74%) and 16 (84.21%) respectively. In PP population, no patient achieved response or complete response. All individuals were defined “censored” at the end of observation.

Response Duration (RD)

The analyses of response duration were performed only on those patients who achieved the type of response specified in each particular statistics. Relatively to ITT population, duration of overall complete response was thus evaluated only on the single patient who reached confirmed CR during the study period.

This patient did not present with worsening of his status during the observation period, no failures were observed relatively to this endpoint. Similarly, duration of overall response applied only to patients whose best overall response was CR or PR. Out of the total of these individuals, two (66.67%) had worsening of their conditions constituting failure events, while the single subject remaining (33.33%) was considered as censored. Duration of stable disease applied only to patients that reached CR, PR or SD, i.e. to 12 individuals out of 20 enrolled, seven of which (58.33%) experienced death or progression of disease; five (41.67%) were classified as censored.

Relatively to PP population, no patient achieved CR or PR, thus duration of overall complete response and duration of overall response were not computed. Five patients were classified as stable disease: 3 out of them (60%) experienced one of the events considered for failure; the remaining 2 (40%) were considered as censored.

Safety

The percentage of patients who experienced one or more adverse event was 90% (18 subjects out of 20). In terms of SOC, many adverse events were classified as “Blood and lymphatic system disorders” (17 patients), “General disorders and administration site conditions” (11 subjects) and “Vascular disorders” (8 patients). Fourteen patients had adverse events of toxicity grade 3-4, for a total of 33 severe adverse events. A great number of events were considered as related to chemotherapy, including events related to haematological toxicity in a large number of patients, 70% experiencing thrombocytopenia and 60% experiencing neutropenia, gastrointestinal disorders in 20% of patients and others. A lower number of events was considered to be related to Avastin, including hypertension in 40% of patients and bleeding in various organs districts in 15% of patients. Events considered as related to both Avastin and chemotherapy included asthenia in 35% of patients and one case of pulmonary embolism.

Four patients complained of serious adverse events, three of which drug related (one thrombocytopenia related to fotemustine, the other two cases, metrorrhagia and pulmonary embolism, related to both fotemustine and Avastin). The last one serious adverse event led to the discontinuation of the treatment.

During the study 12 patients out of 20 (60%) died. All deaths observed were due to disease progression. No important differences were observed in time profiles of weight, body surface area, pulse rate and blood pressure, although a minimal increase was noted in systolic and diastolic values, and a decrease was observed for weight and body surface area. Anyway, the number of patients observed during the last visits was very small, making these changes unreliable. All observed patients reported an ECOG Performance status equal to 0 (fully active, able to carry out all pre-disease performance without restriction) at all time points.

Conclusions:

In conclusion, the combination of Avastin and fotemustine had an unfavourable safety profile that led to the discontinuation from the treatment of the majority of the enrolled patients, without the sought benefits from the combination therapy on disease control.

Date of report:

12.04.2011

Click here for the protocol registry listing of this trial.