94IN - Perspectives of subtyping in soft tissue sarcomas: The relevance for clinical trial design

Date 29 September 2012
Event ESMO Congress 2012
Session Subtyping soft tissue sarcomas for treatment approaches
Topics Pathology/Molecular Biology
Soft Tissue Sarcomas
Basic Scientific Principles
Presenter Jaap Verweij
Authors J. Verweij
  • Dept. Of Medical Oncology, Erasmus University Medical Center - Daniel den Hoed Cancer Center, 3015CE - Rotterdam/NL


In the relatively recent past soft tissue sarcomas were lumped together for the purpose of studying treatment effects in clinical trials. This has clearly served a purpose in the development of non-specific cytotoxic agents. But even for cytotoxic agents more recently hints of selective sensitivity of sarcoma subtypes emerged. So one can state in general that using unselected populations in soft tissue sarcoma trials increases the ease of patient accrual, but at the same time diminishes the chance of achieving a statistically significant clinical benefit. The identification of KIT as a single tumor growth driver in Gastrointestinal stromal tumor (GIST), and its subsequent effective inhibition with imatinib, have further strengthened the realization that the various subtypes of soft tissue sarcomas should each be considered as unique diseases. Because of the rarity of soft tissue sarcomas in general and the linked limited number of patients for each individual sarcoma subtype, the now required further subsetting creates a challenge for current and future trial design. Evidence from recent trials suggests that, if they are based on well-characterized preclinical data and/or clinical observations, performance of trials in select histologic subtypes should be pursued and is feasible; at least for tumor cell related targeted therapeutic approaches. For targets in the tumor environment there are also signals that subsets may be more sensitive, although this does not ensure that the other subsets are totally insensitive. There seems to be a gradual differentiation of sensitivity. For several agents the sensitivity of subsets could be predicted with predictive biomarkers, but biomarkers were of more limited value of biomarkers in assessing individual clinical benefit. This yields yet another challenge, certainly for agents where growth inhibition rather than tumor regression is expected. All of these aspects are guiding us towards new trial approaches that will shortly be discussed.


The author has declared no conflicts of interest.