Targeted therapies in the treatment of advanced basal cell carcinoma

Objective: To review available treatment options for the management of advanced basal cell carcinoma (BCC) with a focus on hedgehog inhibitor therapies. Methods: A MEDLINE search of hedgehog inhibitors in the treatment of advanced BCC was conducted. Results: BCC is the most commonly diagnosed skin cancer worldwide. While the majority of BCC can be effectively treated with standard surgical excision, Mohs surgery, curettage and electrodessication, radiotherapy, and/or superficial field therapies, rarely, patients develop advanced BCC (ie, locally advanced or metastatic BCC), which can be associated with treatment challenges and poor outcomes. Among several therapies being investigated for the treatment of advanced BCC, hedgehog pathway inhibitors have emerged as an important treatment option for this population. Vismodegib and sonidegib are currently the only hedgehog pathway inhibitors approved for the treatment of certain subsets of patients with advanced BCC. Clinical efficacy and safety data from the pivotal phase 2 clinical trials of vismodegib and sonidegib in patients with advanced BCC are reviewed. Data on other hedgehog inhibitors (eg, taladegib, patidegib, itraconazole, posaconazole) are also presented. Because resistance to hedgehog inhibitors may occur, use of novel combinations and/or immunotherapy strategies are also being evaluated. Conclusions: Hedgehog pathway inhibitors have an established efficacy and safety profile and have become an important treatment option for patients with advanced BCC. The optimal use of hedgehog pathway inhibitors in these patients continues to evolve. *Correspondence to: Michael R. Migden, Departments of Dermatology and Head and Neck Surgery, MD Anderson Cancer Center Houston, TX USA, Tel: 713-500-8260; Fax 713-745-3597; E-mail: mrmigden@mdanderson.org


Introduction
Basal cell carcinoma (BCC) is one of the most common malignancies among Caucasians [1]. BCC is typically a slowly growing, locally invasive tumor, with a low metastatic rate (less than 0.5%) [2,3], although its incidence has been increasing [1]. Overexposure to ultraviolet radiation from the sun is an important risk factor for the development of BCC [1,4]. Other risk factors for BCC include increasing age, male sex, prior arsenic exposure, fair skin type, and immunosuppression [5].
Most cases of BCC are treatable if caught sufficiently early [3]. Infrequently, patients with BCC present with lesions not amenable to surgery or radiation therapy and/or lesions that are locally advanced, invasive, or metastatic; these lesions are referred to by some as advanced BCC [2,4]. In addition, locally advanced BCC (laBCC) can be classified by histology as aggressive or nonaggressive. Aggressive histologies range from less aggressive subtypes (such as nodular BCC) to those of more aggressive subtypes (morpheaform, infiltrated, sclerosing, or basosquamous) [6]. Other factors may affect the complexity of BCC and subsequent management decisions, such as lesion size and location: a lesion amenable to surgery on the trunk may be more difficult to treat on the face [4]. Furthermore, challenges associated with the management of BCC include number of lesions and recurrence rates. Cure rates for recurrent lesions are lower than for primary lesions, and the risk of recurrence may remain high even with additional surgery, especially for lesions on the face. Finally, the presence of genodermatoses affects disease management, as these syndromes give rise to multiple and recurrent lesions. Genodermatoses include BCC nevus syndrome (BCCNS), xeroderma pigmentosum, Bazex-Dupré-Christol syndrome, and Rombo syndrome.

Management of advanced BCC
Treatment options for BCC depend on lesion type and location, and patient preference, and include surgical excision, Mohs surgery, curettage and electrodessication, radiation therapy, topical treatments (eg, topical 5-fluorouracil, imiquimod), photodynamic therapy, cryotherapy, and systemic drug treatment (Table 1) [7,8]. Each treatment modality has advantages and disadvantages, as well as clinical indications for which it is best suited. Of the options listed in Table 1, hedgehog pathway inhibitors (HHIs) are the sole systemic drug treatment for advanced BCC.
The hedgehog (HH) pathway is involved in organ formation during embryogenesis. Normally, the HH pathway is dormant in adulthood, but it may be reactivated during conditions in which adult replacement tissue is created or physiologically necessary (eg, taste buds, hair follicles, and wound repair) [9]. It may also be aberrantly activated in several types of cancer. Ligand activation of the cell surface molecule Patched-1 (Ptch1) releases inhibition of Smoothened (Smo), thereby allowing its activation. Smo translocates to and accumulates in the primary cilia, culminating in activation of transcription factors Gli1, Gli2, and Gli3 [9][10][11]. The relationship between HH signaling and human cancer was first noted in patients with BCCNS who had inactivating mutations in Ptch1, causing constitutive activation of HH signaling in the absence of ligand [11,12]. In BCC, up to 75% of cases show mutations in Ptch1, indicating a strong role for aberrant HH signaling in the disease [11,13].
Cyclopamine, a teratogen found in corn lily plants, was the first discovered HH pathway antagonist [14,15]. Cyclopamine binds to Smo, disrupting normal embryonic development. Furthermore, in mouse xenografts, cyclopamine inhibited proliferation and invasion of glioma, melanoma, and colon, pancreatic, and prostate cancers in orthotopic models [16]. Modifications of cyclopamine created a semisynthetic HHI, patidegib, with greater potency and better pharmacokinetics [17]. The discovery of cyclopamine demonstrated that small molecules could inhibit HH. Therefore, the search for more HHIs was undertaken using high-throughput screening, leading to the discovery of vismodegib and sonidegib [18][19][20]. Using a Gli1 reporter gene assay, itraconazole was discovered to inhibit HH signaling [21,22]. Itraconazole acts by preventing Smo translocalization and accumulation in primary cilia [21]. Vismodegib and sonidegib have been approved in the United States (US) and European Union for the treatment of laBCC; in Australia and Switzerland, both agents are approved treatments for laBCC and metastatic BCC (mBCC) [23][24][25][26][27][28][29].
Because the HH pathway may be activated downstream of Smo, researchers are examining the possibility of inhibiting Gli-mediated transcription directly. To this end, the compounds GANT-58 and GANT-61 were found to suppress Gli-mediated signaling in a reporter gene assay [30]. In tumor xenografts, GANT-61 was found to inhibit proliferation and induce apoptosis [31]. These agents are not yet in clinical trials.
STEVIE was a single-arm, open-label, multicenter, post-approval trial in 1215 patients (1119 laBCC, 96 mBCC) receiving vismodegib 150 mg/day [34,35]. The primary end point of this trial was safety, and end points in STEVIE were evaluated only by investigators (Table 2). At 12 months, 12% of patients were still receiving vismodegib, with TEAEs being the main reason for discontinuation of therapy. The majority of

Sonidegib
In the multicenter, randomized, double-blind, pivotal phase 2 BOLT trial, which resulted in approval of the 200-mg daily dose, responses to treatment were evaluated differently than in ERIVANCE or STEVIE, in that both central and investigator assessments were used throughout the 42-month trial duration (Table 2) [32][33][34][35][36][37]. The primary end point of BOLT was ORR, consisting of those patients who had either CR or PR, per central review [36]. In BOLT, patients' lesions were evaluated by BCC-modified RECIST (BCC-mRECIST), which is more stringent than the RECIST used in ERIVANCE and STEVIE [38]. BCC-mRECIST is a multimodal tumor assessment method integrating magnetic resonance imaging per RECIST v1.1, standard and annotated color photography per World Health Organization (WHO) guidelines, and histology in multiple biopsy specimens surveying the lesion area [38]. A prespecified sensitivity analysis using an alternative definition for CR (at least a PR according to MRI and/or photography and no evidence of tumor on biopsy of the residual lesion) yielded a CR rate of 21%, similar to that reported in ERIVANCE [39,40].
A total of 230 patients (laBCC, n=194; mBCC, n=36) were treated with sonidegib (200 mg or 800 mg) once per day [36,40]. In patients with laBCC, at 12 months per central review, ORRs (CR + PR) were 57.6% and 43.8% in the 200-mg and 800-mg arms, respectively. The mDOR (central review) was not reached in patients who received 200 mg but was 15.7 months in the 800-mg arm. In patients with mBCC, ORRs were 7.7% for the 200-mg arm and 17.4% for the 800-mg arm. The mDOR was not reached in either group of patients (200-mg or 800-mg). Investigator-reported response rates in patients with laBCC and mBCC were higher than those determined via central review.

Other hedgehog inhibitors for basal cell carcinoma
Taladegib (LY2940680) is a novel Smo antagonist in early stage clinical trials for advanced BCC and other solid tumors [42]. In a phase 1 trial of 84 patients with laBCC or mBCC (NCT01226485), responses were observed in patients, whether HHI-naive (11 of 16 patients) or previously treated with an HHI (11 of 31). Four patients had events considered dose limiting.
As previously mentioned, itraconazole inhibits the translocation and accumulation of Smo in the primary cilium [21,22]. In an exploratory phase 2 trial of 29 patients with BCC (each of whom had at least one lesion greater than 4 mm in diameter), itraconazole was effective: of 19 treated patients, four achieved a PR and four achieved stable disease (SD) [44]. Further, on average, in vismodegib-naive patients, itraconazole reduced tumor area by 24% (95% CI, 18.2%-30.0%).
Posaconazole is a second-generation triazole that, like itraconazole, inhibits the HH pathway [45]. Posaconazole demonstrated activity against Smo-mutant cell lines and synergized with vismodegib in HHdependent models. It is not yet in clinical trials for efficacy in cancer treatment.

Resistance to hedgehog inhibitors
Of patients who failed to respond to vismodegib in ERIVANCE, it is possible that failure was due in part to spontaneous mutations in Smo [46]. Research has identified several mutations in Smo giving rise to HHI resistance [47,48]. For example, a mutation at residue 518 of Smo increased binding affinity for sonidegib and decreased binding affinity for vismodegib [49,50]. Also, the D427H mutation of Smo disrupted binding between sonidegib and Smo protein and changed the conformation of its transmembrane domain [51]. The D473H mutation alters the conformation of Smo such that the HHIs vismodegib and sonidegib can no longer bind with sufficient affinity to inhibit Smo activity. Using X-ray crystallographs to model changes in Smo, a molecule with HHI activity that would not lose activity in the presence of the D473H mutation (LEQ-506) was developed. LEQ-506 has not yet been tested for efficacy.
Smo mutations were identified in 22 of 44 clinical specimens resistant to HHIs obtained from patients with BCC [52]. In a case report, resistance to vismodegib developed in a patient with BCC; genomic analysis revealed that the recurrent lesions had two different mutations in Smo. In some instances, multiple mutations causing resistance have been observed within the same tumor, as identified by genomic analysis [53]. Resistance was observed in two patients with laBCC or BCCNS who were treated with vismodegib, which was discontinued to manage AEs. When the drug was restarted, however, efficacy was lost [54]. To examine retreatment with HHIs in detail, a set of 6 patients enrolled in STEVIE who discontinued vismodegib due to progressive disease were tried on a second course of vismodegib [55]. The ORR was 80% on the first course and 50% for the second course in this set of patients. The authors concluded that retreatment with the same HHI is feasible and that the same drug may have activity after initial disease progression.
One patient with laBCC resistant to vismodegib responded to a combination of sonidegib and itraconazole [56]. This case is especially noteworthy because the patient presented with intracranial, inoperable advanced BCC involving the sinuses and brain. Itraconazole may have additional antineoplastic effects, such as angiogenesis inhibition and inducing cell cycle arrest [22]. Itraconazole plus arsenic trioxide also induced SD in 3 of 4 patients with mBCC who completed 3 cycles of drug [57].
The role of the phosphoinositide 3-kinase (PI3K) pathway in promoting resistance to HHIs has also been studied [58]. In a proof of concept study in smoothened inhibitor-resistant patients with advanced BCC, the majority (5 of 7 evaluable patients) had either a PR (n=1) or SD (n=4) when given sonidegib plus the PI3K inhibitor buparlisib [59]. However, this study was terminated early due to toxicity of the combination.
Besides combining distinct HHIs or an HHI with a drug having a different mechanism of action, it may be feasible to target HH components downstream of Smo, such as Gli1/2 as described above with GANT-61 [59]. Exploratory research in this area is underway, as is work on other compounds that inhibit Smo through mechanisms differing from those of currently approved HHIs [60,61].

Conclusions and future directions
As HHIs continue to be used for advanced BCC, the question of overcoming resistance to them will be of increasing importance. In addition to targeting mediators downstream of Smo and using novel combinations of drugs that suppress more than one signaling pathway, using the body's immune system may also be an emerging treatment option. Anti-programmed death-1 (PD-1) immunotherapy was successfully used for a patient with mBCC who could not tolerate HHIs [62]. Furthermore, the use of HHIs as neoadjuvant therapy is under investigation [63]. There remains much to be learned regarding the optimal use of HHIs in BCC.

Source of economic support
Publication of this review was supported by Sun Pharmaceutical Industries, Ltd.