Selectively Targeting Mutant BRAF in Thyroid Cancer

Salerno et al. (1) in this issue of the Journal present the first preclinical thyroid cancer studies to use a promising category of targeted therapy, drugs that selectively inhibit the mutant BRAF oncoprotein. BRAF has occupied a central role in thyroid cancer pathogenesis since the original identification of BRAF mutations in papillary and anaplastic thyroid cancer (2,3). Numerous follow-up studies have shown an adverse impact on prognosis of the predominant V600E mutation (4,5). This molecular lesion is present in about 40% of papillary and 25% of anaplastic lesions and usually appears without other simultaneous molecular changes in the canonical ras-RAF-MAPK kinase (MEK)-ERK pathway (3). BRAF mutations correlate strongly with radioiodine resistance and reduced expression of the sodium iodide symporter (6).

In 2006 Solit et al. (7) published a landmark preclinical study in melanoma and colorectal cancer, showing that the presence of BRAF mutations correlated with growth arrest, induced by inhibitors of the MEK kinase, which acts downstream of BRAF. Cell lines lacking BRAF mutations were resistant to the MEK inhibitor, with the exception of ras mutant lines, which were variably sensitive. These data suggested that BRAF mutant tumors might be uniquely dependent on growth signaling through the RAF-MEK-ERK pathway. Several studies have subsequently extended these preclinical findings in thyroid cancer, showing the cytostatic action of MEK inhibitors in BRAF mutant thyroid cancer cell lines (8,9). On a parallel track, promising phase II clinical data have indicated activity of sorafenib in patients with differentiated thyroid cancer (10). To this day, the contribution of RAF inhibition to sorafenib’s antitumor activity remains uncertain, although antiangiogenic actions via inhibition of vascular endothelial growth factor receptor may be more significant than RAF inhibition by this drug.

Several pharmaceutical companies have pursued the strategy of developing small molecule inhibitors of BRAF that would be selective for the predominant V600E mutant form. Malignant melanoma, with about 9000 deaths per year in the United States and a high prevalence of BRAF mutations, has been the main focus of this drug development effort. A phase I clinical trial of the selective BRAF inhibitor PLX4032 presented at the American Society of Clinical Oncology Annual Meeting in May 2009 showed that five of the seven BRAF V600E melanoma patients treated at effective dose levels had significant tumor regression (11). Because phase I studies are more focused on dose exploration than efficacy, this preliminary indication of activity is quite encouraging. The three thyroid cancer patients with BRAF mutations included in this trial also had at least minor responses to the drug (11). Additional drugs are being developed that would target mutant BRAF with differing selectivity including RAF265/Chir-265, PLX4720, and XL281 (12).

In this rapidly moving field, Salerno et al. (1) now report the first detailed preclinical data for selective BRAF inhibitors in thyroid cancer cells. One of the two tested drugs, PLX4032, exhibits greater than 10-fold selectivity for the V600E form of BRAF compared with wild-type BRAF and no significant activity in vitro against a panel of other tested kinases (1,12). As anticipated, PLX4032 caused down-regulation of MEK and ERK kinase activity, specifically in thyroid cancer cell lines harboring the V600E BRAF mutation. Normal thyrocytes and cell lines bearing ras and RET/PTC1 mutations were remarkably unaffected by the drug. In parallel, PLX4032 potently inhibited growth in BRAF mutant cell lines with no effect on normal thyrocytes or thyroid cancer cells with other mutations. Similar to data for MEK inhibitors, a G₁ cell cycle arrest was observed with no evidence for cell death.

In a fascinating twist, inhibition of ERK (and to a lesser extent, MEK) was incomplete in the BRAF mutant cell lines, even at high doses of PLX4032. These results are in sharp contrast to the MEK inhibitor PD0325901, which fully inhibited ERK activation. The authors postulated that disabled feedback mechanisms including ERK phosphatases might be responsible for higher levels of active ERK in the setting of the selective BRAF inhibitor. Alternatively, signaling through wild-type RAF proteins could play a role in this phenomenon. Regardless of the mechanism, it is uncertain whether residual signaling through the RAF-MEK-ERK pathway could mitigate the activity of these drugs. An intriguing possibility not explored in this work is that mutant BRAF could be acting through other signaling intermediates besides MEK and ERK, including nuclear factor-κB (13). It will be important to learn whether inhibition of these additional signaling pathways by BRAF-selective drugs could be relevant in thyroid cancer or melanoma. A disappointing feature of the study by Salerno et al. (1) is the relatively limited data for activity in nude mouse xenograft tumors. The authors reported results only extending to 11 d for a single drug dose, showing a trend to smaller tumor volumes that did not reach statistical significance by ANOVA.

Whether selective BRAF inhibitors will eventually find a role in thyroid cancer therapy is difficult to predict. An appealing feature of these findings is the striking selectivity for the mutant form of BRAF. Increasingly, a patient’s tumor genotype will come to influence therapeutic decision making in advanced thyroid cancer. Lessons learned from the highly selective clinical activity of drugs such as trastuzamab (Herceptin) in patients with Her2/Neu amplified breast cancer, and erlotinib (Tarceva) in epidermal growth factor receptor mutated lung cancer will help to inform clinical trials design in thyroid cancer. Indeed, it seems quite apparent from this study and recent targeted therapy clinical trials that future trials of BRAF inhibitors in thyroid cancer should be focused on the patient population with BRAF mutations, rather than a broader population including patients with other ras pathway mutations.