Possible explanations
The designs of the phase 1b/2 and phase 3 studies had some important differences. The phase 1b/2 study was a small, open-label, US-centric study (10 sites) that did not include a placebo or subtype- specified analyses. Its primary endpoint was PFS, it did not have a loading dose of olaratumab, and it specified the timing of dexrazoxane administration after 300 mg/m2 of doxorubicin.
ANNOUNCE, on the other hand, was a large (n=509), international (110 study sites), double-blind, placebo-controlled trial that had outcomes evaluated in STS and leiomyosarcoma. Its primary endpoint was OS, it had a loading dose of olaratumab of 20 mg/kg, and there was no restriction as to the timing of dexrazoxane administration.
Dr. Tap pointed out that in ANNOUNCE it was difficult to predict or control for factors that may have had an unanticipated influence on outcomes, such as albumin levels as a surrogate for disease burden and behavior of PDGFRα status. It is possible, he said, that olaratumab has no activity in STS and that the phase 1b/2 results were due to, among other things, the small sample size, numerous represented histologies with disparate clinical behavior, and the effect of subtype-specific therapies on overall survival, given subsequently or even by chance. On the other hand, it is also possible, he said, that olaratumab has some activity in STS, with outcomes being affected by the heterogeneity of the study populations, differences in trial design, and the performance of the ANNOUNCE control arm. Whatever the case, he said, accelerated approval allowed patients to have access to a potentially life-prolonging drug with little added toxicity.
Discussion
In the expert discussion following the presentation, Jaap Verweij, MD, PhD, of Erasmus University Medical Center in Rotterdam, The Netherlands, congratulated the investigators for performing the study at an unprecedented pace. He commented that lumping STS subtypes together is problematic, as different histological subtypes behave as though they are different diseases. Small numbers of each tumor subtype and subtypes with slow tumor growth can impact trial outcomes. In the phase 1b/2 and phase 3 trials, 26 different subtypes were represented in each study. Dr. Verweij pointed out this could have made a big difference in the phase 1b/2 study, in which there were only 66 patients in each arm.
It is striking to note, he said, that without exception, phase 2 randomized studies in STS involving doxorubicin consistently overestimated and wrongly predicted PFS in the subsequent phase 3 studies. And the situation is similar for OS. The results of the ANNOUNCE study are no exception, he added. “Taken together, these studies indicate that phase 2 studies in soft tissue sarcomas, certainly those involving additions of drugs to doxorubicin, even if randomized, should be interpreted with great caution,” he said.
SOURCE: Tap WD, et al. J Clin Oncol 37, 2019 (suppl; abstr LBA3)
The study was sponsored by Eli Lilly and Company.
Dr. Tap reported research funding from Lilly and Dr. Verweij had nothing to report related to this study. Abstract coauthors disclosed numerous financial relationships, including consulting/advisory roles and/or research funding from Lilly, and several were employed by Lilly.
Addition of Temozolomide May Improve Outcomes in RMS
Investigators from the European Pediatric Soft Tissue Sarcoma Study Group (EpSSG) found that the addition of temozolomide (T) to vincristine and irinotecan (VI) may improve outcomes in adults and children with relapsed or refractory rhabdomyosarcoma (RMS). Principal investigator of the study, Anne Sophie Defachelles, MD, pediatric oncologist at the Centre Oscar Lambret in Lille, France, presented the results on behalf of the EpSSG.