Led by George Demetri, MD, of Dana-Farber Cancer Institute, Boston, expert medical oncology investigators from leading academic sarcoma centers and Janssen Pharmaceuticals employees performed a phase III trial in order to provide evidence for the U.S. Food and Drug Administration (FDA) approval of trabectedin (Yondelis).1,2 This approval was long overdue. The phase III study is summarized in this issue of The ASCO Post.
In an editorial that accompanied the phase III trial publication,3 Gary Schwartz, MD, of Herbert Irving Comprehensive Cancer Center, New York, opined that the study confirmed “what our European colleagues claimed a decade ago.” In fact, at the 2009 ASCO Annual Meeting, the American community of sarcoma physicians was nearly unanimous that trabectedin was an active drug useful for prolonged periods of time for a minority of sarcoma patients. The challenge was to design a convincing strategy to demonstrate this.
Noteworthy Benefit for Some
Based in large part upon this trial, the FDA made the right decision to approve this drug. Trabectedin has modest activity against some soft-tissue sarcoma histotypes (myxoid round cell liposarcoma, leiomyosarcoma). For those patients who benefit, the beneficial period is noteworthy. The toxicity profile for both trabectedin and imidazole carboxamide (or dacarbazine) is comparable and usually easy to manage.
What was unusual about the regulatory agency’s decision was that the study’s primary endpoint of overall survival was not met. Instead, the study showed an increase in the median progression-free survival of 4.2 months compared to 1.5 months in the control group treated with dacarbazine.
Again, the trial raises the age-old question of how to define clinical benefit. Oncologists—sarcoma oncologists, in particular—continue to grapple with this question. This is perhaps in large part because there are so few drugs to treat metastatic sarcoma with unambiguous benefit (think imatinib).
Study Design Details
The study demonstrated the superiority of trabectedin over dacarbazine using a 2:1 randomization design. Randomization gives credence to the comparison of progression-free survival. Too often, particularly in sarcoma trials, a statement is made about progression-free survival in a single-arm study and compared to the European Organisation for Treatment of Cancer (EORTC) idea4 that progression-free survival can be a principal endpoint in phase II trials. However, without demonstrating that the cancer was progressing immediately before receiving a new therapy, progression-free survival can be misestimated by a few weeks or months. Randomization alone gives protection against a false observation of a progression-free survival difference.
Using a 2:1 schema led more patients to receive the “experimental drug” than the approved standard. In turn, this likely reduced the total number of patients required for statistical certainty of an observation. Compared to a 1:1 schema, Korn and Freidlin5 estimated that a 2:1 schema increases the total number of patients by 15% while maintaining the same power calculation.
To some extent, however, the reduced number of patients receiving dacarbazine compromised the comparison of liposarcoma subtypes. There were 47 patients in the liposarcoma control arm, and only 3 patients were reported to have the pleomorphic liposarcoma subtype. The median reported difference in progression-free survival is quite small in dedifferentiated and pleomorphic liposarcoma subtypes. Median progression-free survival with dacarbazine was 1.9 and 1.4 months for dedifferentiated and pleomorphic subtypes, compared with 2.2 and 1.5 months with trabectedin. Thus, the effect on progression-free survival overall in liposarcoma was 1.5 months for dacarbazine and 3.0 months for trabectedin. This difference was largely driven by the effect in the myxoid/round cell variant.
This study included an independent imaging audit in a subset of 60% of the total population. The limited audit confirmed the investigator interpretation of the clinical course of patients in the trial. Perhaps it is time to end the practice of an independent imaging review. This practice was put in place to address the potential bias of the radiologist interpreting the clinical scans. There was the potential for bias when the radiologist would review the patient’s chart to see which treatment had been received. It is an ongoing challenge to have the radiologist first review the clinical record. Instead, the funding should be used to support pathology review. In sarcoma, the incidence of histologic subtype error varies by as much as 25%.
Mechanism of Action
Approximately 40% of all sarcomas have an associated chromosomal translocation and subsequent fusion gene that acts as a transcriptional regulator. This has been known and accepted for decades, yet no direct therapeutic intervention has been forthcoming despite knowing the molecular target. Is trabectedin the first drug to directly interfere with the fusion protein FUS-CHOP that results from t(12;16) (q13;p11), the balanced translocation pathognomonic of myxoid/round cell liposarcoma?
DiGiandomenico et al6 provided xenograft-derived data that suggest a selective mechanism of action of trabectedin in the myxoid/round cell variant, whereby it causes functional deactivation of the chimera and subsequent derepression of differentiation. If this is so, it raises some questions: Is there another mechanism to explain the occasional patient with another subtype of liposarcoma that benefits from trabectedin? Should other soft-tissue sarcomas, such as synovial sarcoma, have been included in this study rather than excluded? Should bone sarcomas such as Ewing sarcoma, which also has a pathogenomic fusion gene to target, have been included? Again, preliminary data from Grohar et al7 suggested this possibility.
Hopefully, with additional laboratory and clinical studies, these and other questions can be addressed. Many have commented that the real drug development work does not begin until regulatory approval. ■
Disclosure: Dr. Baker is a member of the data monitoring committee for Morphotek and is on the advisory committee for Teva.
References
1. FDA approves trabectedin for advanced liposarcoma and leiomyosarcoma. The ASCO Post, October 23, 2015.
2. Demetri GD, von Mehren M, Jones RL, et al: Efficacy and safety of trabectedin or dacarbazine for metastatic liposarcoma or leiomyosarcoma after failure of conventional chemotherapy. J Clin Oncol. September 14, 2015 (early release online).
3. Schwartz GK: Trabectedin and the L-sarcomas: A decade-long odyssey. J Clin Oncol. September 14, 2015 (early release online).
4. Van Glabbeke M, Verweij J, Judson I, et al: Progression-free rate as the principal end-point for phase II trials in soft-tissue sarcomas. Eur J Cancer 38:543-549, 2002.
5. Korn EL, Freidlin B: Outcome-adaptive randomization: Is it useful? J Clin Oncol 29:771-776, 2011.
6. DiGiandomenico S, Frapolli R, Bello E, et al: Mode of action of trabectedin in myxoid liposarcomas. Oncogene 33:5201-5210, 2014.
7. Grohar PJ, Segars LE, Yeung C, et al: Dual targeting of EWS-FLI1 activity and the associated DNA damage response with trabectedin and SN38 synergistically inhibits Ewing sarcoma cell growth. Clin Cancer Res 20:1190-1203, 2014.
