Pharmacokinetics Pharmacokinetic analysis was carried out by MDS Pharma Services (2350 Cohen Street, St Laurent, Quebec, Canada), using HPLC combined with triple mass spectrometric detection. Pharmacokinetic parameters for erlotinib were evaluated in the first cycle with erlotinib, in patients in cohort 1 only. Blood samples were collected into lithium heparin tubes pre-dosing and thorough at 0.5, 1, 2, 3, 4, 6, 8, 10 and 24h post-dosing on days ?1, 1 and 7. Plasma samples, prepared by centrifugation within 1h of collection, were stored frozen at ?70��C until analysis. The following PK parameters were evaluated for erlotinib: maximum observed concentration (Cmax), time to peak concentration (Tmax) and area under the curve (AUC 0�C24h). These parameters were assessed on the day prior to chemotherapy, on the day of chemotherapy and 6 days after chemotherapy.

Antitumour efficacy At baseline, each patient underwent a pelvic examination and an abdomino-pelvic computed tomography (CT) scan, and blood levels of the ovarian tumour marker CA-125 were assessed. Patients with palpable lesions had a pelvic examination before each cycle. After cycles 3 and 6, abdomino-pelvic CT scans were performed on patients who had (a) disease evident at baseline and (b) a negative baseline scan with CA-125 evidence of disease progression. CA-125 blood levels were measured before each treatment cycle and during follow-up. Evaluation of tumour response was based on the Response Evaluation Criteria in Solid Tumours (RECIST) and on changes in CA-125 levels. Progression-free survival was defined as time to progression or death from any cause.

Progressive disease was defined as either clinical evidence of progressive disease based on either the RECIST Response Criteria for Solid Tumours or elevated CA125 levels as defined as either (a) an increase by twice the upper limit of normal for patients where the Ca-125 normalises or is never elevelated, or (b) an increase by twice the nadir value for patients with elevated pretreatment and who do not normalise values after therapy (any elevated levels have to be confirmed by repeat testing). Statistical methods In this study, a planned cohort size of 12 evaluable patients in each dose escalation cohort was used, rather than the usual six. The main reason for this was to reduce the risk of the MTD being established on the basis of docetaxel/carboplatin toxicities alone.

In addition, the larger sample size provided greater power to detect any potential effect of erlotinib on nadir neutrophil counts in cohort 1. The cohort of the dose level to be taken forward to phase III was planned to be expanded to 20 patients, to allow more experience with the combination at this dose level. For cohort 1, nadir neutrophil Drug_discovery data and erlotinib PK were analysed using repeated measures analysis of variance (ANOVA) techniques.