Treatment of Recurrent Ewing Sarcoma

The rEEcur trial for Ewing Sarcoma

Obligatory—This is not medical advice

You must love the lengths investigators and sponsors go to produce a catchy name. The study title, fully composed, is ‘An international randomized controlled trial of chemotherapy for the treatment of recurrent and primary refractory Ewing sarcoma.’ I prefer saying rEECur trial, though, so I get it. This study was performed primarily in Europe. The protocol is at this footnote.¹ The design of this all was fairly complicated, and I’ll do my best to explain it. In essence, this is a design that sought to perform an initial efficacy evaluation for response to therapy (in the Phase II portion), which eventually led into a Phase III run-off evaluating event free survival (EFS). Event free survival is something we call a composite outcome. It includes multiple components and any of them happening signifies an event. These may include:

1. Time of progression by RECIST v1.1 (growth of tumor in sum of greatest dimensions by >20%)

2. Recurrence of disease following response (either shrinkage of disease followed by regrowth by 20%, or reappearance of measurable disease after complete response)

3. Second malignancy (new cancer)àdistinguishes this endpoint from progression free survival

4. Death (of any cause)

Do I think that event free survival is the perfect endpoint? Not necessarily, but I’m a little bit of a zealot for overall survival as the purest means of analyzing efficacy for cancer care. Arguments against using overall survival include time to that outcome. In aggressive disease states, like recurrent Ewing Sarcoma, this is not necessarily as true.²

In general, utilizing overall survival in lieu of progression free survival may only lead to an approximately 12% change in the length of the drug development cycle (and this is an average—it’s shorter for more rapidly fatal conditions). The 5-year survival for relapsed Ewing sarcoma is only 15%, indicating that overall survival could have been assessed without using too much more time. In fact, planning for the phase III portion had anticipated years would need to pass to get the statistical evidence proposed: 6.25 years for 3 arms, and 4.2 years for 2 arms. This all would be contingent upon rate of accrual, anticipated survival after therapy, etc. OS would not have been too different. Apologies for not doing the calculations, but I’m strapped for time today.

Eligibility

Moving on to the eligibility criteria. I find these all reasonable. You need to have Ewings. You need to be fit enough to get treatment. The age range is also appropriate for a largely pediatric disease. Patients were randomized to one of 4 arms initially. Response rate after 4 cycles was done to determine whether each arm would be eliminated (rate of reduction in sum of largest diameters by >30%). Response to therapy has been an indicator of PFS and OS in some smaller series.³ This, however, has only been done in a small group of patients analyzed with severe limitations. They saw that patients with more responsive tumors lived longer, not that regimens which cause more responses are more effective (eg help people live longer and feel better).

The other interesting aspect to this is that the investigators seem to admit that this is somewhat of an odd means of assessment within their protocol. See page 8 of the protocol in reference #1. They were unable to find an association at their timepoints between OS and EFS, they subsequently incorporated secondary endpoints to determine which arms would move forward. The success of each agent was charted over time and reported at ASCO 2019 ⁴, ASCO 2020⁵, and ASCO 2022⁶

The short story, as is shown here, is that in sequence, gemcitabine/docetaxel, then irinotecan/temozolomide were eliminated in the Phase II portion (see ASCO 2019 and 2020 abstracts). This culiminated in a phase III runoff for event free survival.

From rEECur protocol (ref #1)

The final ASCO presentation, which was a plenary, and is reference #6 as per above showed an increase in EFS (explained above) for patients treated with high dose ifosfamide (3.7 mo vs. 5.7 mo with overlapping CI). There was also an increase in OS, as a secondary endpoint, at 10.4 months for topotecan/cytoxan vs. 16.8 months for high dose ifosfamide.

What do we do with these results?

Larger questions remain about the best way to incorporate this data into practice. Within the protocol, there is an allowance that patients need to be able to be randomized between multiple arms, but stipulates that perhaps not all arms would be available at all sites. The implication of this is that more intensive and/or expensive therapies may not be available at all participating medical centers. Those medical centers with the capacity to deliver these medications are likely to also have stronger support in other areas and have more resources. Therefore, one may wonder if this may selectively improve EFS for those regimens being administered at centers with more resources.

I don’t know how much I use these data—particularly as we continue to perform new trials of medications for Ewing sarcoma, such as TKI or other targeted treatments.⁷ I think of the rEECur study as one that perhaps adds to the landscape of cytotoxic treatment for Ewing sarcoma, but could have been more. This is not untrue of other phase III trials conducted in recent years for rare disease states. The results are only as good or meaningful as the design that produced them.

1. https://www.birmingham.ac.uk/research/crctu/trials/reecur/index.aspx

2. https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2729389

3. https://ascopubs.org/doi/10.1200/JCO.2005.05.105

4. https://ascopubs.org/doi/abs/10.1200/JCO.2019.37.15_suppl.11007

5. https://ascopubs.org/doi/abs/10.1200/JCO.2020.38.15_suppl.11502

6. https://ascopubs.org/doi/abs/10.1200/JCO.2022.40.17_suppl.LBA2

7. https://www.thelancet.com/journals/lanonc/article/PIIS1470-2045(19)30825-3/fulltext