Experience with independent radiological review during a topotecan trial in ovarian cancer

The results of phase II clinical trials are usually based on response of tumours to new oncolytic agents as evidenced by radiological imaging techniques. In this trial, all claimed responders were reviewed at a specially convened meeting by the peer group of study investigators and a radiologist, independent of the study institutions.

One hundred eleven patients with advanced epithelial ovarian cancer who had previously been treated with a platinum based regimen and had subsequently relapsed and who had measurable disease were treated with topotecan at a dose of 1.5 mg/m2/day i.v. on five consecutive days repeated every 21 days to assess efficacy and tolerability. Ninety-three were considered eligible for the study per protocol and lesions were assessed by either computerised tomography (CT) or ultrasound (US). At the meeting, scans from all 24 (25.8%) claimed responders were reviewed, lesions remeasured by the radiologist and a group discussion led to a final response classification.

Ninety-two patients were found to be eligible for the study and 14 (15.2%) were confirmed as responders. Ten were rejected as responders, mainly because the lesion did not decrease in size by < or = 50%, but one patient failed to meet the entry criteria. Remeasurement of CT scans was more objective than US scans. Difficulties were encountered during review of some CT scan sequences because of non-uniform imaging parameters.

Independent radiological review in conjunction with the peer review group in this trial enabled rigorous and consistent application of response criteria. This decreased the response rate from 25.8% to 15.2%, but this represents a more objective assessment. CT scanning is an objective technique for assessing response rates in phase II studies whereas US is subjective and dose not necessarily allow accurate lesion assessment on subsequent examinations, nor allows independent review at a later date. For these reasons it should not be used in such studies for accurate lesion assessment. Cross-sectional imaging techniques such as CT and magnetic resonance imaging (MRI) do allow accurate lesion assessment and independent review at a later date, but standard protocols need to be instituted, to allow consistency and a comparison to be made with subsequent studies using the same agent and a broad comparison to be made with other agents.