Targeted Therapy for Advanced Basal-Cell Carcinoma: Vismodegib and Beyond

The Hedgehog pathway plays a critical role in embryonic development and is not active in most adult tissues, with the exception of stem cells, hair follicles, and skin cells in which the pathway is important for cell maintenance [16]. Key components of the Hedgehog pathway were first described in 1980 by Nusslein-Volhard et al. [17]. Hedgehog pathway signaling starts with the Hedgehog (Hh) ligand binding to a 12-pass transmembrane receptor known as PTCH, which is located in the primary cilium of the cell (Fig. 1). In 1993, three mammalian homologs of the Hh ligand were described, including Sonic Hedgehog (named after the popular Sega videogame), and also Indian Hedgehog and Desert Hedgehog (both named after real hedgehog species) [18, 19]. The “Hedgehog” moniker was coined based on the description of Drosophila melanogaster (fruit fly) larvae, which took on the appearance of the spiky hedgehog when the gene was mutated.

In the absence of the Hh ligand, the PTCH receptor acts as a tumor suppressor by inhibiting the next protein in the pathway known as Smoothened (Smo), which is a G-protein-coupled receptor. When the Hh ligand binds to PTCH, the inhibitory effects on Smo are released allowing the signal to propagate. Although the mechanisms following Smo inhibition release have not been completely elucidated, Smo activation ultimately results in the release of inhibition of glioma-associated protein (Gli) through the suppressor of fused molecule (Sufu). The Gli family of proteins (Gli-1–3) are zinc finger transcription factors that are capable of activating a number of target genes, which can result in an oncogenic effect on the cell. Among the genes that are upregulated through Gli transcriptional activation are PTCH1 (provides negative feedback of pathway), Gli-1 (positive pathway feedback), and other gene pathways that aid in the survival of the cell, such as angiogenesis [20, 21], cell cycle regulation [22], and antiapoptosis pathways [23]. The Hedgehog pathway also conducts significant crosstalk with other molecules and pathways including p53, Wnt, PI3 K/aKT/mTOR, and retinoic acid. These interactions create a complex network, which may promote a variety of resistance mechanisms for drug targeting of this pathway. A variety of diseases have been linked to abnormal Hedgehog pathway signaling besides BCC, including medulloblastoma, hematologic malignancies, and other solid tumors [24].

Cyclopamine, the first naturally occurring inhibitor of the Hedgehog pathway, was isolated from the Veratrum californicum (the California corn lily) plant [25]. Cyclopamine was named after its ability to induce cyclopia and holoprosencephaly in the progeny of animals that fed on the plant while pregnant, highlighting the alkaloid’s role in impairing the Hedgehog pathway in developing embryos. Cyclopamine was first found to bind to the Smo receptor resulting in blockade of downstream Hh signaling pathway transduction [26]. This discovery has led to a variety of more potent and selective Smo antagonists, which have been developed and incorporated into clinical research for a variety of cancer types, including patients with advanced BCC.

A PubMed search was utilized to retrieve the data presented in this review article. The search terms BCC, metastatic basal-cell carcinoma, smoothened inhibitor, Hedgehog pathway, vismodegib, LDE-225, and targeted therapy were used for this search. Additionally, abstracts from national cancer meetings were obtained via a similar search on the American Society of Clinical Oncology (ASCO) website and the European Society of Medical Oncology (ESMO) website.

Vismodegib (GDC-0449) is a first-in-class, orally bioavailable inhibitor of Smo. Based on the efficacy and tolerability results of recent clinical studies, vismodegib received US Food and Drug Administration (FDA) approval on January 30, 2012. Table 1 [27‐32] summarizes the critical studies that have explored the use of vismodegib for BCC.

The first study was published in 2009 by Von Hoff et al. [27] and described the results of a dose-escalation phase 1 trial of vismodegib in patients with metastatic or locally advanced BCC. In this study, 68 solid tumor patients, including 33 patients with metastatic or locally advanced BCC, were treated with three different doses of drug. The study consisted of two phases, a dose-escalation phase followed by an expansion cohort. In the dose escalation phase of the trial, patients received 150, 270, or 540 mg daily, with each dosing group including one BCC. Based on pharmacokinetic studies, 150 mg daily was chosen as the optimal dose from this first stage of the trial. The second stage involved expansion cohorts, including 12 non-BCC patients, 20 patients with advanced BCC (dosing included 150 or 270 mg/day), and 16 patients with solid tumors (10 of which had advanced BCC). Results of the complete study, including other solid tumor types, have been presented elsewhere and demonstrated activity in medulloblastoma in addition to BCC [31]. For those with BCC, key inclusion requirements included histologically confirmed locally advanced or metastatic BCC considered refractory to standard therapy, Eastern Cooperative Oncology Group (ECOG) performance status 0–2, absence of prolonged QT interval, and negative pregnancy test for females. Patients who had radiographically measurable disease had imaging studies at baseline and every 8 weeks with responses measured by Response Evaluation Criteria In Solid Tumors (RECIST) (version 1.0) criteria.

Of the 33 patients enrolled with BCC, 17 patients received GDC-0449 at a dose of 150 mg daily, 15 received 270 mg daily, and one patient received a dose of 540 mg daily. Additionally, 15 patients (45%) had locally advanced BCC and 18 patients (55%) had metastatic BCC. Prior treatments, included surgery (n = 28, 85%), radiation therapy (n = 19, 58%), and systemic therapy (n = 15, 45%). Of the 18 patients with metastatic BCC, there were 15 with radiographically measurable disease, with seven of these having a partial response (>30% shrinkage of tumor size). Two additional patients had partial responses based on imaging and physical exam. Seven patients had stable disease and two patients had progressive disease as their best response. The overall response rate (ORR) was 50% for metastatic BCC. Of the 15 patients with locally advanced BCC, there were two complete responses, seven patients with partial response, four patients with stable disease, and two patients with progressive disease as best response. The ORR for locally advanced BCC was 60%. At the time of study publication, the median duration of response was 8.8 months and ongoing. In terms of toxicities, there were no dose-limiting adverse effects noted. There was an isolated case of grade 4 asymptomatic hyponatremia. Grade 3 events included fatigue, hyponatremia, weight loss, dyspnea, muscle spasm, and prolonged QT interval. Correlative molecular studies in select patients showed mutations in the PTCH1 gene in nine of ten specimens analyzed, loss of heterozygosity of the PTCH1 gene in one patient’s tumor, SMO-M2 mutations in one patient, PTCH1 mutations in normal skin of a patient with basal-cell nevus syndrome, and elevated Gli-1 mRNA expression in 25 of 26 biopsied tumors. Of the four patients with progressive disease, one patient was not found to have Hedgehog pathway signaling, whereas two of the others analyzed did have signaling suggesting other more important driving mutations for these patients.

Via cell-based high-throughput screening, LDE225 was identified as a selective, potent inhibitor of Smo, which is currently in clinical development for BCC and other cancer types [31]. It has been evaluated in a phase 1 dose-escalation study, which was reported at ASCO in 2010 [32]. In this study, patients with advanced solid tumors were treated with varied doses of LDE225 (100, 200, 400, and 800 mg) with the primary goal of determining the maximally tolerated dose (MTD). At the time of the presentation, 25 patients had been treated with no dose-limiting toxicities noted. Common side-effects included fatigue, nausea, vomiting, anorexia, muscle cramps, and dysgeusia. Additionally, at the time of the report one patient with medulloblastoma had obtained an objective response, whereas five other patients had at least stable disease for 4 months (including one BCC patient). A reduction in Gli-1 mRNA expression was noted in skin samples of patients with the level of reduction correlating with dose of LDE225.

LDE225 has subsequently entered testing in a phase 2 randomized trial for patients with locally advanced or metastatic BCC in 2011, which is currently ongoing. In this study, patients are randomized to one of two different doses of LDE225. The primary endpoint of the study is objective response rate by 6 months with secondary endpoints including duration of response, PFS, and safety. Additionally, a trial of LDE225 in patients who have progressed on another Smo-inhibitor (e.g., vismodegib) is currently being conducted, which will shed light on the role of cross-resistance with these inhibitors (ClinicalTrials.gov number: NCT01529450).

LDE225 has also been evaluated in the management of nevoid BCC syndrome. In this trial, eight patients were treated with 0.75% LDE225 and vehicle [33]. Lesions on each patient were randomized to the active LDE225 cream or vehicle only. There were a total of 27 lesions treated with either the LDE225 cream or vehicle twice daily for 4 weeks. Fourteen lesions received vehicle whereas 13 lesions received LDE225 cream. There were clinical responses in all lesions treated with LDE225 except one. Clinical responses included three complete clinical responses and nine partial responses. Clinical response was seen in only one of the 14 lesions treated with the vehicle. Both the LDE225 cream and vehicle were well tolerated with no reported skin toxicities. Systemic absorption of LDE225 was below the limit of detection in 50% of the patients with the highest concentration noted in the other four patients being 0.11 ng/mL. Correlative analysis on biopsied tumors after LDE225 topical treatment revealed downregulation of Hedgehog pathway gene targets, such as Gli-1, Gli-2, PTCH1, and PTCH2. Certainly, for patients with numerous localized BCCs, such as in nevoid BCC syndrome, the use of topical LDE225 appears to be a rational approach as it avoids systemic exposure that predisposes to more side-effects.

A variety of other agents that inhibit Smo are in clinical development. These agents have been evaluated, or are currently being evaluated, in phase 1 clinical studies. These include BMS-833923, which has been described in a phase 1 clinical trial in advanced solid tumors [34]. This study was reported in 2010 at ASCO, and at the time of the report 18 patients had been treated with BMS-833923 at doses ranging from 30 to 240 mg. One patient with Gorlin syndrome who was treated with the 240 mg dose had a confirmed partial response, and another patient with medulloblastoma had stable disease lasting for more than 11 months. Further updates from this trial are expected. Currently, trials with BMS-833923 are ongoing in small cell lung cancer, chronic myeloid leukemia (CML), multiple myeloma, and gastrointestinal malignancies. Several other Smo inhibitors are being evaluated in first-in-human clinical studies. This includes TAK-441 (Millennium, Cambridge, MA, USA; ClinicalTrials.gov number: NCT01204073) and LEQ506 (Novartis, Basel, Switzerland; ClinicalTrials.gov number: NCT01106508), which are also in early phase studies with advanced solid tumors.

Although treatment of BCC with vismodegib and similar agents has resulted in dramatic responses, resistance to Smo inhibition occurs resulting in new tumor development or growth of previously responding tumors. A recent analysis into the mechanism of resistance to vismodegib has recently shed light into the complex nature of this process [35]. In a patient with medulloblastoma with a known PTCH mutation, the initial response to vismodegib was seen with subsequent progression of disease [36]. Comparison of before treatment tumor samples and samples obtained from a vismodegib-resistant tumor demonstrated a novel finding. The pre-existing PTCH mutation found in the pre-treatment tumor was still present in the resistant tumor; however, a new mutation in SMO (D473) was seen. Similar to alterations in BCR–ABL, which confer resistance of CML cells to imatinib, this SMO-D473 mutation was found to confer resistance to vismodegib. Further evaluation implicated that this particular mutation affected vismodegib binding to Smo. A panel evaluating other point mutations in this particular location uncovered potential oncogenic properties with autoactivation of Smo signaling [37]. Evaluation of other Smo inhibiting agents, demonstrated several potential candidates that could inhibit wild-type Smo and this new D473 mutation. Thus development of next-generation Smo inhibitors, which have a broader Smo inhibitory profile, could be a key to unlocking more durable benefits. Additionally, inhibition of downstream molecules, such as Gli, could have benefits to patients with Smo-inhibitor refractory tumors or as vertical inhibition combination strategies with current Smo inhibitors. Several Gli-inhibitor molecules have been discovered, including GANT 58 and GANT 61; however, agents such as these have yet to be evaluated in clinical studies [38]. Finally, other pathways play important roles via either downstream Hedgehog pathway regulation or Hedgehog pathway crosstalk. Some BCC cell lines that demonstrate Smo-inhibitor resistance rely on the PI3 K-Akt pathway. Coinhibition of Smo and PI3 kinase in these Smo-inhibitor resistant cell lines has demonstrated subsequent tumor responsiveness to treatment, implicating possible horizontal pathway inhibition strategies, which could be employed clinically [39]. Although, the mechanisms of resistance to Smo inhibitors remains incompletely evaluated, further information is likely to be forthcoming as ongoing trials are collecting tumor samples upon progression, which should further guide future treatment development strategies.

Vismodegib represents a first-in-class medication for use in patients with locally advanced or metastatic BCC. As with any new drug, there is a learning curve to overcome in the community. Before the clinical development and approval of vismodegib, there was little thought of using systemic therapy in these patients as chemotherapy has little activity. Advanced BCC patients are cared for by a wide range of disciplines including dermatologists, Mohs surgeons, plastic surgeons, otolaryngologists, and radiation oncologists. With the approval of vismodegib, medical oncologists now have a larger role to play. Common questions include: Who is the appropriate patient to receive this agent? How long should treatment with vismodegib be continued in someone with a complete response? What is done for patients who do not tolerate the medication? What do we do upon progression? Certainly, some of these questions can be answered with evidence-based support; however, others remain the focus of ongoing and future research.

The introduction of selective, potent inhibitors of the Hedgehog pathway has led to improved outcomes for patients with advanced BCC who previously had limited systemic treatment options. Vismodegib represents a first-in-class Smo inhibitor, which has shown prominent clinical activity in phase 2 trials for patients with advanced BCC and Gorlin syndrome. Unfortunately, most patients treated with this drug eventually have disease progression. This is anticipated for other Smo inhibitors in development as well. Therefore, it is critical that further research to help our understanding of Smo inhibitor resistance be performed. Many of the trials that are ongoing are actively collecting samples from progressing lesions for molecular analysis. It is anticipated that more than one mechanism of resistance will be identified. Early evaluation of one patient who has progressed on vismodegib has revealed that mutation in the Smo molecule can occur, which interferes with vismodegib binding. This mutation is quite similar to the mutations noted in BCR–ABL, which develop in response to exposure to imatinib. This finding opens the door for the evaluation of agents that bind to both wild-type and mutant SMO as a means of overcoming vismodegib resistance. Additionally, agents that target downstream molecules in the Hedgehog pathway, such as Gli or other pathways, which contribute to Hedgehog pathway inhibitor resistance, such as the PI3kinase pathway are also candidates for overcoming resistance. Certainly, these are areas that need to be further explored as new agents that have similar activities are introduced. Finally, it is important to evaluate these agents earlier in the disease process as potential adjuvant or neoadjuvant adjuncts to traditional approaches, which may result in better outcomes and hopefully prevent the devastating occurrences of locally advanced and metastatic forms of BCC.

The identification of the Hedgehog pathway’s role in BCC, as well as drugs that are able to target this pathway, has led to a critical proof-of-concept translation of these agents into the clinical management of advanced BCC. The first-in-class Smo inhibitor, vismodegib, has given the clinician an important tool in treating patients with this devastating disease. It is critically important that physicians understand when and how to use this novel agent in the management of these patients. Other agents that work similarly to vismodegib are in development and are expected to expand the clinical options for these patients even further. Research into mechanisms of resistance of Smo inhibitors, identification of other relevant molecular targets and an understanding of the use of Hedgehog pathway inhibitors in earlier stage disease remains a crucial next step to improving outcomes for patients BCC.