Paul­ B.­ Chapman,­ MD­

Attending Physician Melanoma/Sarcoma Service
Memorial Sloan-Kettering Cancer Center Professor of Medicine
Weill Cornell University Medical College New York City

The overall median survival for patients with stage IV melanoma is less than a year, with approximately 10 percent long-term survivors. Although patients can respond to chemotherapy, it has not been possible to show that it improves median overall survival. High-dose interleukin-2 (IL2), an FDA- approved treatment for stage IV, can induce long-term responses in a small percentage of patients (<10 percent) but is highly toxic and difficult to administer. The discovery in 2002 that about half of melanomas harbor an activating mutation in the BRAF gene1 provided a novel target for therapy which has led, in a relatively short period of time, to a new treatment paradigm for metastatic melanoma.

One of the primary evolutionary problems a eukaryotic cell had to overcome was how to transmit extra-cellular signals from the cell surface to the nucleus. One of the primary path- ways that evolved was the MAP kinase (MAPK) pathway (Figure 1a). When a ligand binds to the appropriate receptor tyrosine kinase, a conformational change is induced that leads to the phosphorylation and activation of Ras (H, N, and K). Activated Ras induces the dimerization of the Raf kinases (A, B, and C) as well as their phosphorylation. These activated dimers phosphorylate MEK, which phosphorylates ERK. Activated ERK enters the nucleus and acts as a transcription factor, turning on the transcription of several genes that lead to cellular proliferation and survival. There is also a complicated negative feedback mechanism that keeps this system under control.

Approximately 50 percent of melanomas harbor an activating mutation in the BRAF gene that leads to an activating mutation of the valine at position 600, usually to glutamic acid (V600E) but sometimes to arginine or lysine. In these cells, the MAPK pathway is constitutively activated, being driven by this mutated BRAF kinase. These cells do not have activated Ras, so the Raf kinases are not dimerized. This is an important mechanism for the specificity of BRAF inhibitors.

Vemurafenib (also known as PLX4032), binds to the ATP-binding site of mutated BRAF and locks it into an active conformation, but without ATP, the mutated BRAF cannot phosphorylate downstream MEK and the MAPK pathway is turned off.

However, vemurafenib can also bind to CRAF and to a lesser degree, wild- type BRAF.2 In the wild-type cell, the Raf kinases are dimerized, and when vemurafenib binds to wild-type CRAF or BRAF, this conformational change induces a conformational change in the other member of the dimer pair resulting in transactivation and increased MAPK pathway activity.3 Thus, the specificity of vemurafenib comes from the fact that in cells with BRAF mutations, the cell is driven by activated BRAF, which exists primarily as monomers and is inhibited by the drug (Figure 1b). This model is supported by data from several in vitro studies.

The first human phase I trial with vemurafenib (then designated PLX4032) began in 2006. Early on in the study, at the lower dose levels, it became clear that the drug could induce dramatic responses in melanoma tumors harboring a V600E mutation4 (Figure 2). The maximum tolerated dose (and the recommended phase II dose) was deter- mined to be 960 mg po bid. An extension cohort of 32 melanoma patients with V600E mutations treated at this recommended phase II dose showed that most patients experienced tumor shrinkage and that 56 percent qualified as having a partial or complete response by RECIST [Response Evaluation Criteria in Solid Tumors] criteria. At 2 years, 7 of 22 patients were still on the study drug.

Recently, the results of a phase II trial (termed BRIM2) were reported.5 In this multicenter trial, 132 melanoma patients harboring a BRAFV600E mutation who had disease progression after at least one prior therapy were treated with vemurafenib at 960 mg po bid. In this trial, in which the tumor responses were reviewed by a central, independent panel, the response rate by RECIST criteria was 52 percent, consistent with the phase I experience. The median overall survival had not been reached at 12 months, indicating that responses can be durable.

In the two studies, the toxicities seen were similar. Arthralgias and fatigue were the most common dose- limiting toxicities. Other common toxicities were alopecia, rash, palmar- plantar dysesthesia, and photosensitivity. Although these were generally grade I or II in severity, occasionally they were grade III.

Drugs that inhibit Raf kinases are known to induce non-melanoma skin tumors. Consistent with this, ap- proximately 25 percent of patients on vemurafenib developed skin tumors characterized as verrucae, keratoacanthoma, or squamous cell carcinoma. These tumors were simply excised. There were no instances of meta- static cancers (other than melanoma) or squamous cell carcinoma in other anatomical sites beyond the skin.

This high response rate justified conducting a phase III trial in which stage IV (or unresectable stage III) previously untreated melanoma patients whose tumor harbored a BRAFV600E mutation were randomized 1:1 to either vemurafenib at 960 mg po bid or standard chemotherapy (dacarbazine 800 mg/ m2 iv every 3 weeks).6 The primary endpoints were overall survival and progression-free survival. Between January and December, 2010, 675 patients were accrued to this trial among 104 participating centers worldwide. In January 2011, the planned interim analysis was performed, at which time the independent data safety monitoring committee announced that the primary endpoints of the trial had been met and recommended that patients randomized to dacarbazine be allowed to cross over to vemurafenib.

Despite the very short follow-up time (median 3 months), the vemurafenib group had a 63 percent lower hazard of death and a 74 percent lower hazard of progression compared to the dacarbazine group (Figure 3). Although median overall survival could not be reliably estimated at this first analysis due to the very short median follow- up, 6-month overall survival rates were 84 percent for the vemurafenib group compared to 64 percent for the dacarbazine group. As expected, the response rate in the vemurafenib patients was much higher (48 percent) compared to the dacarbazine patients (5.5 percent). Progression-free survival, a co- primary endpoint, was also significantly superior on the vemurafenib arm. The median time to progression was 5.3 months compared to 1.6 months with dacarbazine.

Based on these data, and supporting data from the phase II trial, the FDA approved vemurafenib (trade name ZelborafTM) for the treatment of meta- static melanoma harboring a BRAFV600E mutation on August 17, 2011.

The clinical trials with vemurafenib consistently show a median time to progression of 5-7 months. This indicates that melanomas can develop resistance to vemurafenib relatively quickly. Therefore, it is critical to understand the mechanism of resistance.

The experience with inhibiting mutated KIT in GIST (gastrointestinal stromal tumors) using imatinib predicted that melanomas developing resistance to vemurafenib would develop a second mutation in BRAF that would prevent binding of the drug. This has not been the case. In some cases of resistance, however, the tumor has developed an activating mutation in upstream NRAS which reactivates the MAPK pathway. Recently, Poulikakos and colleagues have reported a novel resistance mechanism in which a splice variant in the mutated BRAF allele results in the loss of a domain that prevents dimerization.7 As a result, the truncated BRAFV600E kinase readily dimerizes with other RAF kinases despite the lack of upstream RAS activation. By the trans- activation mechanism described above (Figure 1b), these dimers now activate the MAPK pathway in the presence of the drug and result in resistance. Other investigators have described potential resistance mechanisms that can reactivate the MAPK pathway through down- stream events, or mechanisms that activate the parallel PI3K [phosphatidylinositol 3-kinases] pathway.8-10 It re- mains to be determined how frequently these mechanisms play a role in patients developing resistance to vemurafenib. It is likely that other mechanisms re- activating the MAPK pathway will be described in these patients.

Vemurafenib was FDA-approved for use in metastatic melanoma just 9 years after the first report that melanomas often harbor BRAFV600E mutations. This represents a remarkable amount of collaborative work among those in the melanoma community and the pharmaceutical industry. Other BRAF inhibitors are also in development, most notably GSK2118436, which has recently completed a phase III trial. In this study, 200 patients were randomized 3:1 to GSK2118436 vs. dacarbazine, the primary endpoint being progression-free survival. Results are expected in 2012.

Although 90 percent of BRAF mutations are V600E, about 5 percent are V600K; other mutations make up the remaining 5 percent. In the phase II and phase III vemurafenib trials, a small number of patients were found retrospectively to have had a V600K mutation. Forty percent of these patients responded, indicating that melanomas with V600K mutations are also sensitive to vemurafenib. In the future, it will be necessary to test the other less frequent, non-V600E mutations.

The efficacy of BRAF inhibitors on brain metastases is currently un- clear. Despite the fact that neither GSK2118436 nor vemurafenib were predicted to cross the blood-brain bar-ier, the phase I trial of GSK2118436 gave encouraging results. Of 10 patients treated with brain metastases, 8 showed shrinkage of  > 30 percent and several had complete responses in the brain. A formal phase II trial in patients with brain metastases has recently completed accrual with GSK2118436 and another is under way with vemurafenib.

The infrequency of complete responses with vemurafenib and the frequency with which melanomas develop resistance suggest that combination therapies will be needed. There are currently ongoing trials adding MEK inhibitors, ipilimumab (YervoyTM), or bevacizumab to BRAF inhibitors.

Other rational combinations are also under consideration. [See “Combining Forces: Vemurafenib and Ipilimumab To Be Tried Together."]

1. Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene in human cancer. Nature 2002; 417:949-54.
2.  Bollag G, Hirth P, Tsai J, et al. Clinical efficacy of RAF inhibitor needs broad target blockade in BRAF-mutant melanoma. Nature 2010; 467:596-9.
3. Poulikakos PI, Zhang C, Bollag G, Shokat KM, Rosen N. RAF inhibitors transactivate RAF dimers and ERK signalling in cells with wild-type BRAF. Nature 2010; 464:427-30.
4. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition of mutated, activated BRAF in metastatic mela- noma. N Engl J Med 2010; 363:809-19.
5. Ribas A, Kim KB, Schuchter LM, et al. BRIM-2: An open-label, multicenter phase II study of RG7204 (PLX4032) in previously treated patients with BRAFV600E mutation-positive melanoma. In: American Society of Clinical Oncology; 2011 June 4, 2011; Chicago; 2011.
6. Chapman PB, Hauschild A, Robert C, et al. Im- proved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med 2011; 364:2507-16.
7. Poulikakos PI, et al. Nature 2011 (in press).
8. Johannessen CM, Boehm JS, Kim SY, et al. COT drives resistance to RAF inhibition through MAP kinase pathway reactivation. Nature 2010; 468:968-72.
9. Nazarian R, Shi H, Wang Q, et al. Melanomas acquire resistance to B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature 2010; 468:973-7.
10. Villanueva J, Vultur A, Lee JT, et al. Acquired resistance to BRAF inhibitors mediated by a RAF kinase switch in melanoma can be overcome by cotargeting MEK and IGF-1R/PI3K. Cancer Cell 2010; 18:683-95.