The Future of Drug Development? Seeking Evidence of Activity of Novel Drugs in Small Groups of Patients

Abstract

We increasingly recognize that our traditional classification of tumor types on the basis of the organ of origin and microscopy alone is inadequate. Tumors grouped according to histologic classification systems are diverse entities that differ greatly in their molecular characteristics, clinical course, and sensitivity to antitumor agents. Accordingly, many tumor types are currently divided into smaller subgroups on the basis of molecular characteristics that have therapeutic consequences. Examples include melanoma, breast cancer, colorectal cancer, gastric cancer, and non–small-cell lung cancer. Undoubtedly, most tumor types as we currently know them will be redefined into smaller subsets of clinically relevant, molecular entities. Soft tissue sarcomas (STS) are well-known to represent a heterogeneous group of tumors comprising more than 50 histotypes. As a result of their rare occurrence, which was thought to preclude the possibility of performing adequately powered studies in specific histotypes, and despite the recognized heterogeneity in clinical course and known differences in sensitivity to chemotherapy, all subtypes from this group were treated similarly until a decade ago. This started to change with the introduction of imatinib with impressive antitumor activity in GI stomal tumors 2 and later dermatofibrosarcoma protuberans, both harboring molecular targets affected by the agent. The activity of imatinib was not observed in any other STS subtype, which emphasized the relevance of the underlying genetic aberrations as the primary determinant of treatment benefit. The total group of STS involves only 1% of all malignancies, and the subtype of alveolar soft part sarcoma (ASPS) accounts for approximately 1% of all STS, making it an extremely rare tumor type with an estimated annual incidence of 100 new cases in the United States. Although ASPS generally exhibits indolent growth, the majority of patients develop metastases during the course of their disease, most frequently in the lungs and bones. Compared with other STS subtypes, ASPS may also have a greater likelihood of metastasizing to the brain. Given that the limited data available suggest that ASPS is relatively resistant to radiotherapy and conventional chemotherapeutic agents, radical surgery is the only meaningful option. For patients with metastatic disease not amenable for radical resection, the median overall survival is 30 months, which is much longer than that for other STS histotypes, consistent with its relatively indolent course. Importantly, ASPS features a specific translocation [t(X,17)(p11; q25)], which fuses the ASPL gene to the TFE2 transcription factor gene. The presence of this fusion protein results in uncontrolled transcription of TFE3-regulated genes such as c-MET and several proangiogenic factors. In the article accompanying this editorial, Kummar et al describe a single-arm phase II study of cediranib in patients with advanced ASPS not amenable to curative surgery. Cediranib is a tyrosine kinase inhibitor (TKI) that targets all three vascular endothelial growth factor receptors (VEGFRs) at low concentrations. The rationale for this study was based on the overexpression of angiogenesispromoting genes in ASPS, the vascular appearance at histopathologic examination, and promising antitumor activity reported in another VEGFR TKI–related small study and in a phase I study of cediranib. Using RECIST-classified response rate (RR) as a primary end point, a Simon optimal two-stage design was applied, in which an RR of 25% was deemed worthwhile for further exploration. In 43 patients evaluable for response, 15 (35%) experienced a partial response. Importantly, only two patients (5%) experienced progression as best response. The progression-free rate at 26 weeks was 84%, whereas some patients still on treatment had not yet reached this time-point. After a dose reduction from 30 to 20 to 15 mg/day, cediranib had a toxicity profile that was deemed tolerable. In view of the high, unprecedented RR in this rare disease and the tumor regressions seen in almost all patients on the study, it is obvious that cediranib exerts antitumor activity against advanced ASPS. However, the study was not randomized and thus the question can be raised whether the activity seen in this trial is sufficiently convincing for implementation into daily clinical practice. Clearly, randomized trials are the preferred cornerstone for approval and treatment decision making in daily clinical practice. Yet, some of the most successful drugs applied in medical oncology have been implemented on the basis of results from nonrandomized studies. Most recently vismodegib, a Hedgehog pathway inhibitor, received marketing approval in the United States for advanced basal cell carcinoma on the basis of a single-arm study with 96 patients. RRs were 30% and 43% in patients with metastatic (n 33) and locally advanced disease (n 63), respectively, with a median duration of response of 7.6 months in JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 31 NUMBER 18 JUNE 2