ICON8: What the Ovarian Cancer Trial Found

A result that did not travel

In Japan, a trial called JGOG3016 had found something striking. Giving paclitaxel weekly instead of every three weeks, alongside three-weekly carboplatin, improved both progression-free and overall survival in ovarian cancer.

That is a big claim, and a cheap one to act on. The drugs were the same. Only the timetable changed.

ICON8 asked whether the same thing would happen in a mostly European population. The answer was no.

How the trial was built

ICON8 was a phase 3 randomized controlled trial run through the Gynecologic Cancer InterGroup. Between June 6, 2011 and November 28, 2014, it randomly assigned 1,566 women.

All had newly diagnosed epithelial ovarian, fallopian tube or primary peritoneal cancer at FIGO stage IC to IV. They entered either after immediate surgery, or before chemotherapy given ahead of a planned later operation.

There were three arms.

• Group 1, the control: carboplatin plus paclitaxel, both given every 3 weeks.
• Group 2: carboplatin every 3 weeks, plus a smaller weekly dose of paclitaxel.
• Group 3: smaller weekly doses of both carboplatin and paclitaxel.

Our page on carboplatin and paclitaxel for ovarian cancer explains what that combination involves.

The co-primary outcomes were progression-free survival and overall survival. The trial was powered to detect a hazard ratio of 0.75 in progression-free survival, and analyzed by intention to treat.

What it found

By February 2017, 1,018 of the women, or 65 percent, had seen their cancer progress.

The weekly schedules did deliver more drug. Over the whole course, the weekly groups received more total paclitaxel than the three-weekly control group.

It did not translate into benefit. Restricted mean survival time, a more stable measure than the median, was 24.4 months in group 1, 24.9 in group 2 and 25.3 in group 3. That is a gap of under one month across three arms.

Median progression-free survival looked better on paper for the weekly arms, at 17.7, 20.8 and 21.0 months. But neither comparison reached statistical significance: log-rank p was 0.35 for group 2 against group 1, and 0.51 for group 3 against group 1.

Grade 3 or 4 side effects were more common with weekly treatment, though the authors describe them as predominantly uncomplicated. Rates of febrile neutropenia, meaning fever with a low white cell count, and of sensory nerve damage were similar across the three groups.

The number nobody quotes

Completion rates tell their own story.

Seventy-two percent of women in the standard arm finished all six protocol-defined cycles. In the weekly arms it was 60 percent and 63 percent.

Looked at another way, 90, 89 and 85 percent completed six platinum-based cycles. So the platinum got delivered. What slipped was the full weekly schedule.

A regimen that a third of patients cannot complete as written is a different proposition in a clinic than it is on a protocol page.

When to get checked

NCI is direct about ovarian cancer: it may cause no early symptoms, and when symptoms do appear the cancer is often already advanced.

The signs NCI lists are all things that have ordinary explanations most of the time.

• Pain, swelling, or a feeling of pressure in the abdomen or pelvis.
• A sudden or frequent urge to urinate.
• Trouble eating, or feeling full quickly.
• A lump in the pelvic area.
• Gas, bloating or constipation.

NCI's rule for what to do with them is a good one. If they get worse, or do not go away on their own, see a doctor. Persistence is the signal, not severity. Our page on persistent bloating covers how this usually gets noticed.

There is no recommended screening test for ovarian cancer in women at average risk.

The group picture

SEER figures describe the United States population, not any individual.

Five-year relative survival for ovarian cancer is 52.0 percent for cases from 2016 to 2022. By stage it is 91.9 percent while confined to the ovary, 70.1 percent once it reaches nearby lymph nodes, and 31.5 percent once it has spread further.

Only 22 percent are found at that first stage. Fifty-four percent are already distant, which is the highest late-stage share of any common cancer, and the reason the overall figure sits where it does.

An estimated 21,010 new cases and 12,450 deaths are projected for 2026. The median age at diagnosis is 63. Our page on ovarian cancer covers the types and treatment.

What to keep in perspective

• A null result is a result. ICON8 did not show that weekly treatment is worse. It showed that the trial could not detect the benefit it was designed to detect.
• This report covers the progression-free survival endpoint. Overall survival was reported separately, later.
• Trials enroll people who meet specific criteria, so the findings may not describe everyone with this cancer.
• The most useful thing here is not about ovarian cancer at all. A striking result in one population is a hypothesis, not a conclusion, until someone repeats it somewhere else. Our page on trial endpoints explains what these measures do and do not capture.

Sources

• Clamp AR et al., Weekly dose-dense chemotherapy in first-line epithelial ovarian, fallopian tube, or primary peritoneal carcinoma treatment (ICON8), Lancet 2019 — NCBI PubMed record via E-utilities — https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&id=31791688&rettype=abstract&retmode=text
• ClinicalTrials.gov API v2, NCT01654146 — https://clinicaltrials.gov/api/v2/studies/NCT01654146
• NCI PDQ, Ovarian Epithelial, Fallopian Tube, and Primary Peritoneal Cancer Treatment (Patient Version) — https://www.cancer.gov/types/ovarian/patient/ovarian-epithelial-treatment-pdq
• SEER Cancer Stat Facts, Ovarian Cancer — https://seer.cancer.gov/statfacts/html/ovary.html

How this article was prepared

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.

The National Cancer Information Foundation publishes Cancer Explained. This page is for learning. It is not medical advice and does not suggest a test or treatment.

See an error, old source, or unclear wording? Tell us.