MOLECULAR CARCINOGENESIS 54:688–697 (2015)
IN PERSPECTIVE Chemopreventive Opportunities to Control Basal Cell Carcinoma: Current Perspectives Cynthia Tilley,1 Gagan Deep,1,2 and Rajesh Agarwal1,2* 1 2
Department of Pharmaceutical Sciences, Skaggs School of Pharmacy and Pharmaceutical Sciences, Aurora, Colorado University of Colorado Cancer Center, University of Colorado, Aurora, Colorado
Basal cell carcinoma (BCC) is a major health problem with approximately 2.8 million new cases diagnosed each year in the United States. BCC incidences have continued to rise due to lack of effective chemopreventive options. One of the key molecular characteristics of BCC is the sustained activation of hedgehog signaling through inactivating mutations in the tumor suppressor gene patch (Ptch) or activating mutations in Smoothened. In the past, several studies have addressed targeting the activated hedgehog pathway for the treatment and prevention of BCC, although with toxic effects. Other studies have attempted BCC chemoprevention through targeting the promotional phase of the disease especially the inflammatory component. The compounds that have been utilized in pre-clinical and/or clinical studies include green and black tea, difluoromethylornithine, thymidine dinucleotide, retinoids, non-steroidal anti-inflammatory drugs, vitamin D3, and silibinin. In this review, we have discussed genetic and epigenetic modifications that occur during BCC development as well as the current state of BCC pre-clinical and clinical chemoprevention studies. © 2015 Wiley Periodicals, Inc. Key words: basal cell carcinoma; chemoprevention; natural products; silibinin; hedgehog signaling; inflammation; ultraviolet B INTRODUCTION There are about 3.5 million new cases of nonmelanoma skin cancer (NMSC) diagnosed in the United States each year corresponding to an overall lifetime risk of one in five people [1]. NMSC are composed mainly of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC); BCC accounts for about 80% of NMSC. Whereas the cure rates for NMSC is up to 99%, treatments are associated with significant morbidity, recurrence, and high cost to the health care system. The national costs of NMSC treatment and management are about $650 million per year [1]. BCC arises in the skin’s basal cells that line the deepest layer of the epidermis. The most important risk factor for BCC is ultraviolet (UV) light exposure which can initiate skin carcinogenesis, although additional evidence has suggested the potential role of several other environmental insults such as exposure to ionizing radiation (IR) and harmful chemicals such as arsenic [2–5]. Additional risk factors for BCC include fair skin, history of sunburns, high cumulative sun exposure, immunosuppression, and genetic diseases [5,6]. Furthermore, BCC risk in organ transplant patients increases by 10-fold [1]. To protect against BCC, standard sun protection measures are useful such as avoiding mid-day sun, seeking shade, wearing sunscreen, and sun protective clothing [4]. Though the association between UV exposure and BCC development is not as well defined as in SCC, given that approximately a third of BCC originates in anatomical sites receiving minimal UV exposure [5]. ß 2015 WILEY PERIODICALS, INC.
Additional evidence suggests BCC to correlate more with intermittent sun exposure versus cumulative sun exposure [2,7]. Besides UV exposure, there is strong evidence from studies of atomic bomb survivors and radiotherapy patients suggesting that exposure to IR can cause BCC [5]; and these findings correlate well with the BCC mouse model where BCCs can be induced through IR exposure in mice heterozygous for the patch (ptch) gene [8]. Furthermore, BCC incidences have continued to rise due to constant increase in longevity as well as relatively higher exposure to UV radiation associated with the depletion of the ozone layer [5]. The enormity of this disease could be gauged from the fact that annual BCC incidences alone are higher than all other cancer incidences combined; therefore, there is an urgent need for effective preventive measures. Chemoprevention is defined as the use of natural, synthetic, or biological agents to prevent, suppress, and reverse the carcinogenic progression [1]. It is ideally effective in the prevention of the disease and should be non-toxic. Candidate patients for skin
Grant sponsor: NCI R01; Grant number: CA140368 *Correspondence to: University of Colorado, Skaggs School of Pharmacy and Pharmaceutical Sciences,12850 E. Montview Blvd, Room V20-2118, Box C238, Aurora, CO 80045. Received 12 March 2015; Revised 9 May 2015; Accepted 15 May 2015 DOI 10.1002/mc.22348 Published online 5 June 2015 in Wiley Online Library (wileyonlinelibrary.com).
CHEMOPREVENTION OF BCC
cancer chemoprevention are patients with a history of NMSC and who are at high risk of developing more NMSC, or are at risk for invasive or metastatic NMSC. Additionally, patients with numerous actinic keratosis (AK), NMSCs in high risk locations, metastatic NMSC, and solid organ transplant recipients are potential candidates for skin cancer chemoprevention. Currently, there are more than 50 chemopreventive agents under phase II development [9], but so far none of the chemopreventive agents are approved for use against BCC. Exposure to UV radiation results in the formation of thymine dimers which if not repaired could lead to fixed mutations and initiation of carcinogenesis [5]. Promotion of the initiated cells to a pre-neoplastic lesion can take up to 10 or more years whereas progression to carcinoma can occur in a year or less. Therefore, the best opportunity for intervention against cancer is during the promotion phase [5]. The primary preventive strategy could be to reduce exposure to carcinogens but since complete avoidance of the sun and other potential carcinogens is not entirely possible, secondary chemopreventive strategies are also important. The secondary strategies are to circumvent the promotion or progression of initiated pre-cancerous cells using non-cytotoxic dietary or non-dietary pharmacologic agents. Herein, we have discussed the genetic and epigenetic factors involved in BCC development and the current state of preclinical and clinical BCC chemoprevention studies. GENETIC AND EPIGENETIC FACTORS IN BCC DEVELOPMENT There are several genetic diseases associated with predisposition to BCC such as Gorlin, Bazex–Dupre– Christol, and Rombo syndrome [3,10]. Gorlin syndrome was first described in 1960s, and is characterized by multiple BCCs, dyskeratotic palmar, and plantar pitting, and odontogenic keratocysts [10,11]. Patients with Gorlin syndrome are born with an inherited mutation in one allele of the ptch gene, similar to sporadic BCC. Ptch1 is part of a receptor complex at the cell surface composed of two transmembrane proteins, Ptch1 and Smoothened (Smo) [12]. Hedgehog (Hh) is a secreted protein that binds to Ptch1 to activate the pathway. Once Hh ligand binds to Ptch1, Smo is relieved from the inhibitory effects of Ptch1 and transduces signal through a series of interacting proteins leading to the activation of Gli (1–3) [12]. Gli1 is an activator of transcriptional machinery while Gli2 and Gli3 could be either activators or suppressors dependent upon cell state and other cellular contextual signals. There is less known about the genetic defects found in Bazex–Dupre–Christol and Rombo syndrome probably due to a small number of people afflicted with these diseases, though the genetic component of Rombo syndrome may be associated with DNA repair and cell cycle [13,14]. There are additional syndromes Molecular Carcinogenesis
689
that are associated with predisposition to both BCC and SCC, such as xeroderma pigmentosum (XP) and albinism [15]. XP is associated with genetic defects in the DNA repair pathway while albinism patients lack melanin, the component for skin pigment [14]. The genetic diseases associated with predisposition to BCC, especially Gorlin syndrome, have immensely helped in current molecular understanding of the disease. Sporadic BCCs are characterized by constitutively activated Hh signaling. This is due to inactivating mutations in ptch gene (70% of cases) or activating mutations in Smo gene (10–20% of cases). A very small number of BCCs also have mutations in ptch2 or sufu gene [10]. BCC seems to originate from the long-term resident progenitor cells of the inter-follicular epidermis and upper fundibulum [16]. They are composed of lobules of basaloids cells with hyperchromatic nuclei and scant cytoplasm [10]. These tumors are locally invasive and rarely metastasize [5,11]. BCC has various morphological growth patterns such as superficial, nodular, micronodular, infiltrating, sclerosing, and fibroepithelial [3]; wherein nodular, micronodular, and superficial are the most common BCC types and are less aggressive than the other BCC subtypes [11,17,18]. Overall, BCC research has been hindered by the inexplicable failure of mutagenic radiations (UV and IR) and chemicals to induce BCC in murine models [3]. Currently, all BCC mouse models target different components of Hh signaling pathway. The model that represents the closest to the human condition is Ptch1 mutant mice [17]. Ptch1 homozygous mice do not live past embryonic day 9.5 due to heart and neural tube closure defects [17,19,20]. However, Ptch1 heterozygous (þ/) mice develop BCC following IR or multiple UVB exposures resulting in the mutation of the remaining intact ptch1 allele [21]. When Ptch1þ/ mice are exposed to UVB three times per week starting at 6 weeks of age, microscopic lesions arise in mice by 6 months in 100% of the mice [2]. At 9 to 12 months of UVB exposure, grossly visible tumors are present wherein 25% are BCC, 25% SCC, and 50% fibrosarcomas [2]. Additional mouse models of BCC include mice overexpressing Hh, oncogenic Smo and Gli1, or Gli2 expression in the skin using keratin 5, 6, or 14 promoters [17]. The involvement of epigenetic changes in BCC tumorigenesis is less clear than the genetic modifications. A study analyzing methylation of the ptch promoter showed infrequent and low-level of methylation in human BCC suggesting it may play a minor role in BCC carcinogenesis [22]. Brinkhuzien et al. [23] investigated the methylation of the signaling related to Sonic Hh (Shh) and Wnt and also suggested their possible involvement in BCC pathobiology. Importantly, they found Shh, adenomatous polyposis coli (APC), secreted frizzled-related protein 5 (SFRP5), and Ras association domain family 1A (RASSFIA) to be more methylated in BCC versus
690
TILLEY ET AL.
normal skin [23]. Another study found similar results in which they analyzed 52 sporadic BCCs and found five genes to show variable frequency of methylation (14–44%), those include DCR2, DCR1, RASSFIA, APC, and death associated-protein kinase (DAPK) promoter [24]. However, another group irradiated human keratinocytes several times with UVB radiation and analyzed the DNA methylation of the irradiated cells versus non-irradiated cells; and found no difference of DNA methylation between the two groups [25]. Overall, the exact contribution of gene methylation in BCC pathogenesis is still unclear, and further studies are warranted in this direction. Promotional Phase of BCC Tumor development requires accumulation of mutations which often develop as a result of repeated carcinogen exposures and/or impaired damage repair mechanism [9]. Cyclobutane pyrimidine dimers (CPDs) are generated after UV exposure and are repaired by nucleotide excision repair (NER) pathway within several days [7,26]. The photoproducts not removed efficiently may lead to mutations in key tumor suppressor genes such as ptch and p53 [21]. There are UVB signature mutations characterized by C to T or CC to TT transitions located at pyrimidine– pyrimidine sequences and are the result of misrepaired CPDs and 6,4-photoproducts (6,4-PPs) [27]. In NMSC, mutations in p53, ras, and ptch are frequently encountered [9]. In BCC, the ptch gene is most commonly mutated although these mutations alone are not sufficient to produce BCCs [9,12,28]. The p53 tumor suppressor gene is also crucial in the development of BCC wherein about 44–100% of patients with BCC have a mutation in p53 [29,30]. In ptch knockout mouse models, deletion of p53 accelerates and enhances BCC tumorigenesis [12,31]. Downstream mediators of Gli have been shown to contribute to BCC tumorigenesis although it is unclear which downstream changes are essential for BCC. The direct downstream targets of Hh signaling are B-cell lymphoma 2 (Bcl2), cyclin D1 (ccnd1), cyclin D2 (ccnd2), forkhead box M1 (FOXM1), forkhead box E1 (FOXE1), Hh interacting protein (Hhip), platelet derived growth factor receptor alpha (PDGFRa), and missing in metastasis (MIM) [2]. Many of these genes are either involved in cell cycle regulation or cell survival and may contribute to the growth and promotion of BCC lesions [2]. Direct Hh target genes that have the consensus Gli binding site are Gli1, Ptch1, Hhip1, Nmyc1, ccnd1, ccnd2, gremlin 1 (Grem1), and follistatin (Fst), wherein some of these genes are involved in the negative feedback loop, others take part in cell proliferation, stemness, and cell survival [32,33]. There are various downstream targets of the Hh signaling pathway, and further understanding of how these molecules contribute to BCC pathogenesis will be crucial for the chemoprevention and treatment of this disease. Molecular Carcinogenesis
In addition to downstream targets of Hh signaling, the status of mouse hair (haired or hairless) can also affect its susceptibility to develop tumors. One recent study has crossed SKH-1 hairless (hr) mice with Ptchþ/ C57Bl/6 mice to generate a Ptchþ/ hr/ mouse [34]. They found enhanced tumor incidence, volume, and number per mouse in the hairless mice versus the haired. This was accompanied by an elevated inflammatory response and expression of mesenchymerelated proteins in the hairless mice. The relative expression levels of Hh signaling genes were not significantly different between groups, suggesting a similar pathogenesis of BCC tumor development [34]. It is of interest to note that the dose, duration, and frequency of UVB irradiation for haired and hairless mice was different; BCCs were induced with a higher and more frequent UVB dose (240 mJ/cm2) three times per week for 44 weeks in haired mice compared with hairless mice that were irradiated with 180 mJ/cm2 twice per week for 32 weeks [34]. This study additionally highlighted the difference in tumor susceptibly between males and females, wherein male mice developed more tumors than females which is similar to what is found in humans [35]. Another important factor that can influence BCC growth is the phase of the hair cycle. One study reported that BCC growth occurs preferentially during the anagen phase, perhaps due to elevated levels of Hh signaling [3]. These factors may be important in humans as well, considering that humans have relatively low level of hair in comparison to mice. BCC Chemoprevention: Pre-Clinical Studies Hh activation is crucial to BCC development; therefore, several Hh inhibitors have been investigated against BCC in order to reduce tumor progression. Systemic Hh signaling inhibition results in serious adverse effects such as muscle spasms, loss of taste, hair loss, fatigue, nausea, and hyponatremia [3]. One way to avoid some of these effects would be to use inhibitors topically to limit systemic exposure, although these topical applications may prove to be less efficacious [31]. One such Hh inhibitor, Cur61414, was shown to be effective in mice but when topically applied to human BCC, it could not clear the tumors and did not reduce Gli1 levels [31]. Another problematic issue with Hh inhibitors is the tumors’ ability to develop resistance over time. GDC0449, a Smo inhibitor, was used to treat a medulloblastoma patient, but slowly the tumor relapsed [36]. Upon investigation, it was discovered that an amino acid substitution in Smo was present which had no effect on the Hh signaling but disrupted the ability of GDC-0449 to bind to Smo [3,36]. Alternative Smo inhibitors are available that target wild type and the mutated portion of Smo, and these could be used to combat the drug resistance [37]. In addition to mutations found in Smo, other mechanisms of drug resistance involve amplification of downstream genes
CHEMOPREVENTION OF BCC
like gli2 and cyclin D1. Therefore, alternative therapeutic and chemoprevention strategies should be investigated to identify better options. In regard to studying the chemoprevention of BCC in pre-clinical models, the Ptch1þ/ mice present an excellent model mimicking the human disease condition. These mice develop similar tumors, have similar abnormalities in Hh activation, and the disease condition is developed in the same tissue (skin), and is induced through the same environmental insults (UV or IR) as in humans [38]. When Ptch1þ/ mice are irradiated with IR, they develop BCC and trichoepitheliomas, and when irradiated with UVB, they develop macroscopic tumors including BCC, SCC, and fibrosarcomas [38]. In Table 1, the BCC chemoprevention pre-clinical studies are summarized. There have been a relatively large number of studies targeting the BCC-induced inflammatory response including the use of non-steroidal anti-inflammatory drugs (NSAIDs) and sulfasalazine (SSZ). NSAIDs inhibit the cyclooxygenase (COX) enzymes (COX-1 and COX-2) responsible for generating prostaglandin mediated inflammation [38]. In one study, Ptch1þ/ mice were fed celecoxib (a specific COX-2 inhibitor) starting at 6 weeks of age, and then BCC was induced through a single dose of IR at 8 weeks of age. Celecoxib effect was not dose-dependent and when combining all doses used, it reduced microscopic BCC burden by 35% [38]. Additionally, the study investigated two other oral NSAIDs that were able to reduce microscopic BCC burden by 50 and 60%, sulindac and MFtricyclic, respectively. SSZ is a nuclear factor of kappa B (NF-kB) inhibitor and significantly decreased tumor number and volume in Ptch1þ/ hr/ mice [34]. Additionally, SSZ treatment decreased the level of Hh signaling components in BCC tumors and surrounding skin. The level of inflammation was also decreased by SSZ treatment, evidenced through decreased expression of p-p65, p-IKK a/b, and COX-2 [34]. There is an extensive amount of research on the chemopreventive effects of tea in SCC models showing promising effects in terms of inhibiting tumor promotion, malignant progression, and targeting various stages of multistep skin carcinogenesis [9]. However, a study conducted in a BCC mouse model testing green and black tea showed no efficacy, and even some higher number of tumors formed in female mice following green and black tea administration [39]. This study clearly highlights the difference in the pathogenesis of BCC and SCC, and also suggests that chemopreventive agents that are effective against SCC might not necessarily have similar efficacy against BCC. In mouse skin carcinogenesis, retinoids target the b-Raf/MAP kinase ERK kinase (MEK)/extracellular regulated MAP kinase (ERK) signaling pathway [4]. Tazarotene, an acetylenic retinoid, has been studied previously against SCC. It is currently approved in the United States for use in conditions such as topical Molecular Carcinogenesis
691
acne, psoriasis, and photoaging. One study showed the anti-BCC efficacy of tazarotene against BCCs induced by both UVB and IR [40]. Topically applied tazarotene inhibited the formation of microscopic BCC and reduced BCC size compared to vehicle. Tazarotene exerts its protective effects through inhibiting the phosphoinositide 3-kinase (PI3K)/ AKT/mammalian target of rapamycin (mTOR) pathway [32]. The importance of this pathway in BCC was evidenced through the use of pathway inhibitors that reduced the proliferation and tumorigenesis of BCC in vitro and in vivo [32]. Another target of interest in BCC is to investigate the DNA repair capacity given that DNA damage could lead to mutations in key molecules. Topical thymidine application as a potential chemopreventive has shown some promise [21]. A study investigated topically applied thymidine dinucleotide (pTT) in UVB irradiated Ptch1þ/ mice. pTT decreased the level of microscopic BCC, CPDs, 8-oxo-20 -dexoyguanosine (8oxo-dG), Ki-67 positive cells, p53 patches, and COX-2 [21]. Additionally, pTT increased the level of apoptosis in UVB-irradiated mice. Overall, this study showed pTT to be a promising chemopreventive drug that could target multiple pathways involved in BCC development. Ornithine decarboxylase (ODC) is another potential molecular target for chemoprevention of BCC. Overexpression of ODC in Ptchþ/ mice resulted in increased UVB-induced tumorigenesis, and pharmacologic inhibition of ODC with alphadifluoromethylornithine (DFMO) reduced tumor number per mouse as well as microscopic BCC area [41]. This study suggested utilizing DFMO as a potential chemopreventive agent against BCC. Vitamin D3 is another agent that has been found to reduce cell proliferation in BCC cells in culture and mice [42]. Ptchþ/ K14-Cre-ER p53fl/fl mice were treated primarily with tamoxifen to delete p53 expression, which accelerates murine BCC carcinogenesis. Mice were then exposed to IR at 8 weeks of age and BCCs formed by 5 to 6 months of age. After BCCs were visible, mice were treated with vitamin D3 for 30 days and tissue changes were visualized through immunohistochemistry. Overall, vitamin D3 treatment decreased the cell proliferation and also reduced the Gli1 mRNA levels [42]. Additionally, a study on calcitriol, the biologically active form of vitamin D3 has shown anti-BCC tumor growth effects in a conditional Ptch mutant mouse model [43]. These are promising results, and demand more studies focusing on the long-term chemopreventive effects of treatment with vitamin D3 and its active form calcitriol against BCC tumors. Silibinin, an active ingredient from the milk thistle plant seed extract, has shown extensive anti-cancer efficacy against various cancer types in animal models [44–51]. Additionally, there is extensive research on silibinin’s anti-cancer effects against SCC in the UVB-induced skin carcinogenesis model,
Molecular Carcinogenesis
2004 [41] 2011 [42] 2011 [43] 2014 [58]
2011 [31] 2010 [38] 2014 [34] 2001 [39] 2004 [40] 2008 [21]
Year (Ref)
Mitogenic signaling, NF-kB p50, c-Fos
VDR, Hh signaling
VDR, Hh signaling
Inhibitor of ODC
DNA repair
PI3K/AKT/mTOR pathway
Pleiotropic
Ptch1þ/ K14-CreER p53 fl/fl mice Ptchflox/flox ERT2þ/ mice Nude mice
Ptchþ/
Ptchþ/
Ptchþ/
Ptchþ/
Ptchþ/ hr/
NF-kB inhibitor
Smo inhibitor COX inhibitors
Mouse model Ptchþ/ K14-CreER2 p53fl/fl mice Ptchþ/ mice
Target
Injection of BCC cells
Tamoxifen
Single dose of IR
UVB 3 times per week
Single dose of IR or UVB 3 times per week UVB 3 times per week
UVB 3 times per week
Diet
Single dose of IR or UVB 3 times per week UVB 2 times per week
Decreased average tumor number and volume per mouse Ineffective
Inhibition of skin Hh signaling and BCC area decreased Reduced microscopic BCC burden
Results
Reduced formation of microscopic BCC Topical Decreased microscopic BCC area and number and level of CPDs, 8-oxo-dG, and Ki-67 were lowered Drinking Reduced number and area of water microscopic BCCs Topical Decreased cell proliferation and Gli1 expression in BCC tumors IP injection Inhibited cell proliferation and showed some decrease in BCC tumor area Oral Reduced tumor volume and inhibition gavage of proliferation as well as NF-kB and AP-1 activation
Drinking water Drinking water Topical
Topical
Treatment type
IR plus tamoxifen injection
Induction of BCC
BCC, basal cell carcinoma; Smo, smoothened; ptch, patch; IR, ionizing radiation; UVB, ultraviolet B; DFMO, difluoromethylornithine; ODC, ornithine decarboxylase; VDR, vitamin D receptor; Hh, hedgehog.
Silibinin
Calcitriol
Vitamin D3
DFMO
Thymidine dinucleotide
Tazarotene
Green and black tea
Celecoxib, Sulindac, MF-tricyclic Sulfasalazine
Cur61414
Agent(s)
Table 1. Summary of BCC Chemoprevention Pre-Clinical Studies
692 TILLEY ET AL.
CHEMOPREVENTION OF BCC
both dietary intake of silibinin as well as topically applied [45,52–57]. Recently, a study from our laboratory has shown that silibinin can inhibit BCC cell growth and clonogenicity in a dose-dependent manner [58]. Additionally, silibinin was shown to induce apoptosis and suggested to work via inhibiting multiple mitogenic signaling pathways (EGFR, ERK, AKT, NF-kB, etc.). In an ectopic allograft model, oral administration of silibinin strongly inhibited BCC cells tumor growth. These initial studies have provided evidence that silibinin can potentially be used as a chemopreventive agent against BCC but additional studies still need to be conducted. BCC Chemoprevention: Clinical Studies There has not been great success in chemoprevention trials of BCC in humans, the various studies summarized herein are depicted in Table 2. In the past, several trials have aimed to analyze the incidences of both SCC and BCC, which can be challenging in itself since the two types of NMSC are so different. One study conducted by the Skin Cancer Prevention study group administered 50 mg dose of b-carotene daily for 5 years to patients with a recent NMSC, and resulted in no significant effect on the incidence of new NMSC [5,59]. Another group looked at the Nutritional Prevention of Cancer study group with 1,312 patients with a history of NMSC. These patients were treated with 200 mg of selenium or placebo for about 4 years with a 6 year follow-up, wherein selenium showed no effect on incidence of BCC or SCC [5,60]. The SW Skin Cancer Prevention study group enrolled 2,297 patients with a history of greater than 10 AK and less than 2 SCC or BCC. They treated the patients with retinol for 5 years and conducted a 3.8 year follow-up but did not observe any effect on BCC incidence [5,61]. Another study conducted in Australia using 30 mg of daily b-carotene and sunscreen showed no effect on the incidence of BCC, though sunscreen use appeared to reduce number of SCC after 4.5 years follow-up [5,62]. These large group studies have thus far shown no effective chemopreventive option for BCC. Studies with retinoids in clinical models have shown a chemopreventive effect of oral retinoids for SCC, but there is not enough evidence for their use to decrease the risk of BCC [1,4,6,63,64]. Retinoids are considered to be natural and synthetic derivatives of vitamin A. They regulate important functions like signaling for growth arrest, differentiation, and apoptosis; additionally, they modify expression of growth factors, signal transduction, extracellular adhesion molecules, and proto-oncogenes [4]. These compounds act through retinoic acid receptors and retinoid x receptors [1,6]. Additionally, the agents must be continued indefinitely to maintain their protective benefits [4]. The side effects and efficacy of oral retinoids are dose-dependent. Side effects include mucocutaneous dryness and irritation, hair loss, hypertriglyceridemia, hypercholesterolemia, liver toxicity, increased intracranial pressure, and Molecular Carcinogenesis
693
skeletal changes [6]. The use of retinoids for chemoprevention is to decrease the number of NMSC to a manageable level. There are two oral retinoids available in the United States, isotretinoin, and acitretin. Acitretin is the primary choice for chemoprevention. Topical retinoids have a response rate of about 50%, therefore, the routine use of topical retinoids for chemoprevention is not recommended. In the Veterans Affairs Topical Tretinoin Chemoprevention Trial (VATTC), patients were administered a high dose of topical tretinoin for 1.5–5.5 years [64]. Tretinoin was ineffective at reducing the risk of new cases of NMSC. Topical tazarotene was tested in a phase II, randomized, double blind, vehicle control study in 34 basal cell nevus syndrome (BCNS) patients, but did not show significant anti-BCC effects [63]. Overall, systemic retinoids appear to have a modest effect on existing tumors and can suppress new tumor development, while topically applied retinoids have not been as effective [6]. Photodynamic therapy (PDT) has been found to reduce the number of new cases of AK but the chemopreventive effect of PDT on NMSC is inconclusive [1]. Cyclic 5-aminolevulinic acid (ALA)-PDT for the prevention of NMSC in high risk organ transplant recipients (OTRs) showed that treatment at 4 to 8 week intervals for 2 years caused significantly fewer incidences of SCC than in the year before treatment [4]. Topical application of methyl aminolevulinate (MAL) and ALA result in the accumulation of protoporphyin IX preferentially in damaged skin lesions. Subsequent irradiation of treated tissue with PDT resulted in high intra-lesional concentrations of reactive oxygen species (ROS) and selective tissue destruction [4]. A case report on a patient with BCNS found MAL-PDT to be an effective chemopreventive agent against new BCC development, though this is a single patient study and needs to be validated in more patients before any conclusions can be drawn from this study [65]. Another agent that could be possibly useful against BCC is 5-fluorouracil (5-FU). 5-FU is approved for treatment of AKs and the 5% cream is approved for superficial BCCs. There is evidence for effectiveness of 5-FU in reducing AKs but it has not been studied as a chemopreventive agent against NMSC. NSAIDs show conflicting results in literature for the chemoprevention of BCC [38,66–69]. A study evaluating data from a clinical practice research database compared prior exposure to systemic NSAIDs between patients with incident BCC and SCC along with patients without NMSC [69]. They found no evidence for reduced risk of BCC in association with systemic NSAIDs use, though long-term use of ibuprofen or aspirin was associated with a small BCC reduction. One randomized, placebo-controlled, double blind trial evaluated the efficacy of celecoxib as a chemopreventive agent for AK in patients with multiple AK; they found after 9 months of treatment, celecoxib
Molecular Carcinogenesis
2014 [69]
2010 [67]
2010 [38]
2010 [70] 2001 [73]
NSAIDs
Celecoxib
Celecoxib
DFMO 30 1,051
Population at very high risk for NMSC
291
60
240
>60,000
1
34
XP
History of prior NMSC
BCNS
10–40 AKs
BCC and SCC diagnosed between 1995 and 2013
BCNS
BCNS
1131
1621
2297
1312
1805
Number of participants Oral administration/5 years
Treatment/duration
Various chemopreventive treatments including MAL-PDT for a few sessions Identified patients with diagnosis of NMSC and assessed prescriptions for systemic NSAIDs Oral administration/9 months
Topically applied/follow-up of 3 years
Topically applied/1.5 to 5.5 years
Oral administration/4.5 years
NSAIDs ineffective at reducing BCC, but reduced BCC risk in regular users of aspirin and ibuprofen Fewer NMSC in the celecoxib arm than in the placebo arm
Effective at reducing new BCCs in single patient
Ineffective
Ineffective
Ineffective
Ineffective
Ineffective
Ineffective
Results
Cohort study design
Analyzed patients taking ACE inhibitors or ARBs/3.4 years
Reduced risks of BCC and SCC in patients taking ACE or ARBs
Phase II-III, double-blind, placebo-controlled, randomized trial Phase II, double-blind, Oral administration/24 months Reduced BCC burden only in patients placebo-controlled, randomized with
IN PERSPECTIVE Chemopreventive Opportunities to Control Basal Cell Carcinoma: Current Perspectives Cynthia Tilley,1 Gagan Deep,1,2 and Rajesh Agarwal1,2* 1 2
Department of Pharmaceutical Sciences, Skaggs School of Pharmacy and Pharmaceutical Sciences, Aurora, Colorado University of Colorado Cancer Center, University of Colorado, Aurora, Colorado
Basal cell carcinoma (BCC) is a major health problem with approximately 2.8 million new cases diagnosed each year in the United States. BCC incidences have continued to rise due to lack of effective chemopreventive options. One of the key molecular characteristics of BCC is the sustained activation of hedgehog signaling through inactivating mutations in the tumor suppressor gene patch (Ptch) or activating mutations in Smoothened. In the past, several studies have addressed targeting the activated hedgehog pathway for the treatment and prevention of BCC, although with toxic effects. Other studies have attempted BCC chemoprevention through targeting the promotional phase of the disease especially the inflammatory component. The compounds that have been utilized in pre-clinical and/or clinical studies include green and black tea, difluoromethylornithine, thymidine dinucleotide, retinoids, non-steroidal anti-inflammatory drugs, vitamin D3, and silibinin. In this review, we have discussed genetic and epigenetic modifications that occur during BCC development as well as the current state of BCC pre-clinical and clinical chemoprevention studies. © 2015 Wiley Periodicals, Inc. Key words: basal cell carcinoma; chemoprevention; natural products; silibinin; hedgehog signaling; inflammation; ultraviolet B INTRODUCTION There are about 3.5 million new cases of nonmelanoma skin cancer (NMSC) diagnosed in the United States each year corresponding to an overall lifetime risk of one in five people [1]. NMSC are composed mainly of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC); BCC accounts for about 80% of NMSC. Whereas the cure rates for NMSC is up to 99%, treatments are associated with significant morbidity, recurrence, and high cost to the health care system. The national costs of NMSC treatment and management are about $650 million per year [1]. BCC arises in the skin’s basal cells that line the deepest layer of the epidermis. The most important risk factor for BCC is ultraviolet (UV) light exposure which can initiate skin carcinogenesis, although additional evidence has suggested the potential role of several other environmental insults such as exposure to ionizing radiation (IR) and harmful chemicals such as arsenic [2–5]. Additional risk factors for BCC include fair skin, history of sunburns, high cumulative sun exposure, immunosuppression, and genetic diseases [5,6]. Furthermore, BCC risk in organ transplant patients increases by 10-fold [1]. To protect against BCC, standard sun protection measures are useful such as avoiding mid-day sun, seeking shade, wearing sunscreen, and sun protective clothing [4]. Though the association between UV exposure and BCC development is not as well defined as in SCC, given that approximately a third of BCC originates in anatomical sites receiving minimal UV exposure [5]. ß 2015 WILEY PERIODICALS, INC.
Additional evidence suggests BCC to correlate more with intermittent sun exposure versus cumulative sun exposure [2,7]. Besides UV exposure, there is strong evidence from studies of atomic bomb survivors and radiotherapy patients suggesting that exposure to IR can cause BCC [5]; and these findings correlate well with the BCC mouse model where BCCs can be induced through IR exposure in mice heterozygous for the patch (ptch) gene [8]. Furthermore, BCC incidences have continued to rise due to constant increase in longevity as well as relatively higher exposure to UV radiation associated with the depletion of the ozone layer [5]. The enormity of this disease could be gauged from the fact that annual BCC incidences alone are higher than all other cancer incidences combined; therefore, there is an urgent need for effective preventive measures. Chemoprevention is defined as the use of natural, synthetic, or biological agents to prevent, suppress, and reverse the carcinogenic progression [1]. It is ideally effective in the prevention of the disease and should be non-toxic. Candidate patients for skin
Grant sponsor: NCI R01; Grant number: CA140368 *Correspondence to: University of Colorado, Skaggs School of Pharmacy and Pharmaceutical Sciences,12850 E. Montview Blvd, Room V20-2118, Box C238, Aurora, CO 80045. Received 12 March 2015; Revised 9 May 2015; Accepted 15 May 2015 DOI 10.1002/mc.22348 Published online 5 June 2015 in Wiley Online Library (wileyonlinelibrary.com).
CHEMOPREVENTION OF BCC
cancer chemoprevention are patients with a history of NMSC and who are at high risk of developing more NMSC, or are at risk for invasive or metastatic NMSC. Additionally, patients with numerous actinic keratosis (AK), NMSCs in high risk locations, metastatic NMSC, and solid organ transplant recipients are potential candidates for skin cancer chemoprevention. Currently, there are more than 50 chemopreventive agents under phase II development [9], but so far none of the chemopreventive agents are approved for use against BCC. Exposure to UV radiation results in the formation of thymine dimers which if not repaired could lead to fixed mutations and initiation of carcinogenesis [5]. Promotion of the initiated cells to a pre-neoplastic lesion can take up to 10 or more years whereas progression to carcinoma can occur in a year or less. Therefore, the best opportunity for intervention against cancer is during the promotion phase [5]. The primary preventive strategy could be to reduce exposure to carcinogens but since complete avoidance of the sun and other potential carcinogens is not entirely possible, secondary chemopreventive strategies are also important. The secondary strategies are to circumvent the promotion or progression of initiated pre-cancerous cells using non-cytotoxic dietary or non-dietary pharmacologic agents. Herein, we have discussed the genetic and epigenetic factors involved in BCC development and the current state of preclinical and clinical BCC chemoprevention studies. GENETIC AND EPIGENETIC FACTORS IN BCC DEVELOPMENT There are several genetic diseases associated with predisposition to BCC such as Gorlin, Bazex–Dupre– Christol, and Rombo syndrome [3,10]. Gorlin syndrome was first described in 1960s, and is characterized by multiple BCCs, dyskeratotic palmar, and plantar pitting, and odontogenic keratocysts [10,11]. Patients with Gorlin syndrome are born with an inherited mutation in one allele of the ptch gene, similar to sporadic BCC. Ptch1 is part of a receptor complex at the cell surface composed of two transmembrane proteins, Ptch1 and Smoothened (Smo) [12]. Hedgehog (Hh) is a secreted protein that binds to Ptch1 to activate the pathway. Once Hh ligand binds to Ptch1, Smo is relieved from the inhibitory effects of Ptch1 and transduces signal through a series of interacting proteins leading to the activation of Gli (1–3) [12]. Gli1 is an activator of transcriptional machinery while Gli2 and Gli3 could be either activators or suppressors dependent upon cell state and other cellular contextual signals. There is less known about the genetic defects found in Bazex–Dupre–Christol and Rombo syndrome probably due to a small number of people afflicted with these diseases, though the genetic component of Rombo syndrome may be associated with DNA repair and cell cycle [13,14]. There are additional syndromes Molecular Carcinogenesis
689
that are associated with predisposition to both BCC and SCC, such as xeroderma pigmentosum (XP) and albinism [15]. XP is associated with genetic defects in the DNA repair pathway while albinism patients lack melanin, the component for skin pigment [14]. The genetic diseases associated with predisposition to BCC, especially Gorlin syndrome, have immensely helped in current molecular understanding of the disease. Sporadic BCCs are characterized by constitutively activated Hh signaling. This is due to inactivating mutations in ptch gene (70% of cases) or activating mutations in Smo gene (10–20% of cases). A very small number of BCCs also have mutations in ptch2 or sufu gene [10]. BCC seems to originate from the long-term resident progenitor cells of the inter-follicular epidermis and upper fundibulum [16]. They are composed of lobules of basaloids cells with hyperchromatic nuclei and scant cytoplasm [10]. These tumors are locally invasive and rarely metastasize [5,11]. BCC has various morphological growth patterns such as superficial, nodular, micronodular, infiltrating, sclerosing, and fibroepithelial [3]; wherein nodular, micronodular, and superficial are the most common BCC types and are less aggressive than the other BCC subtypes [11,17,18]. Overall, BCC research has been hindered by the inexplicable failure of mutagenic radiations (UV and IR) and chemicals to induce BCC in murine models [3]. Currently, all BCC mouse models target different components of Hh signaling pathway. The model that represents the closest to the human condition is Ptch1 mutant mice [17]. Ptch1 homozygous mice do not live past embryonic day 9.5 due to heart and neural tube closure defects [17,19,20]. However, Ptch1 heterozygous (þ/) mice develop BCC following IR or multiple UVB exposures resulting in the mutation of the remaining intact ptch1 allele [21]. When Ptch1þ/ mice are exposed to UVB three times per week starting at 6 weeks of age, microscopic lesions arise in mice by 6 months in 100% of the mice [2]. At 9 to 12 months of UVB exposure, grossly visible tumors are present wherein 25% are BCC, 25% SCC, and 50% fibrosarcomas [2]. Additional mouse models of BCC include mice overexpressing Hh, oncogenic Smo and Gli1, or Gli2 expression in the skin using keratin 5, 6, or 14 promoters [17]. The involvement of epigenetic changes in BCC tumorigenesis is less clear than the genetic modifications. A study analyzing methylation of the ptch promoter showed infrequent and low-level of methylation in human BCC suggesting it may play a minor role in BCC carcinogenesis [22]. Brinkhuzien et al. [23] investigated the methylation of the signaling related to Sonic Hh (Shh) and Wnt and also suggested their possible involvement in BCC pathobiology. Importantly, they found Shh, adenomatous polyposis coli (APC), secreted frizzled-related protein 5 (SFRP5), and Ras association domain family 1A (RASSFIA) to be more methylated in BCC versus
690
TILLEY ET AL.
normal skin [23]. Another study found similar results in which they analyzed 52 sporadic BCCs and found five genes to show variable frequency of methylation (14–44%), those include DCR2, DCR1, RASSFIA, APC, and death associated-protein kinase (DAPK) promoter [24]. However, another group irradiated human keratinocytes several times with UVB radiation and analyzed the DNA methylation of the irradiated cells versus non-irradiated cells; and found no difference of DNA methylation between the two groups [25]. Overall, the exact contribution of gene methylation in BCC pathogenesis is still unclear, and further studies are warranted in this direction. Promotional Phase of BCC Tumor development requires accumulation of mutations which often develop as a result of repeated carcinogen exposures and/or impaired damage repair mechanism [9]. Cyclobutane pyrimidine dimers (CPDs) are generated after UV exposure and are repaired by nucleotide excision repair (NER) pathway within several days [7,26]. The photoproducts not removed efficiently may lead to mutations in key tumor suppressor genes such as ptch and p53 [21]. There are UVB signature mutations characterized by C to T or CC to TT transitions located at pyrimidine– pyrimidine sequences and are the result of misrepaired CPDs and 6,4-photoproducts (6,4-PPs) [27]. In NMSC, mutations in p53, ras, and ptch are frequently encountered [9]. In BCC, the ptch gene is most commonly mutated although these mutations alone are not sufficient to produce BCCs [9,12,28]. The p53 tumor suppressor gene is also crucial in the development of BCC wherein about 44–100% of patients with BCC have a mutation in p53 [29,30]. In ptch knockout mouse models, deletion of p53 accelerates and enhances BCC tumorigenesis [12,31]. Downstream mediators of Gli have been shown to contribute to BCC tumorigenesis although it is unclear which downstream changes are essential for BCC. The direct downstream targets of Hh signaling are B-cell lymphoma 2 (Bcl2), cyclin D1 (ccnd1), cyclin D2 (ccnd2), forkhead box M1 (FOXM1), forkhead box E1 (FOXE1), Hh interacting protein (Hhip), platelet derived growth factor receptor alpha (PDGFRa), and missing in metastasis (MIM) [2]. Many of these genes are either involved in cell cycle regulation or cell survival and may contribute to the growth and promotion of BCC lesions [2]. Direct Hh target genes that have the consensus Gli binding site are Gli1, Ptch1, Hhip1, Nmyc1, ccnd1, ccnd2, gremlin 1 (Grem1), and follistatin (Fst), wherein some of these genes are involved in the negative feedback loop, others take part in cell proliferation, stemness, and cell survival [32,33]. There are various downstream targets of the Hh signaling pathway, and further understanding of how these molecules contribute to BCC pathogenesis will be crucial for the chemoprevention and treatment of this disease. Molecular Carcinogenesis
In addition to downstream targets of Hh signaling, the status of mouse hair (haired or hairless) can also affect its susceptibility to develop tumors. One recent study has crossed SKH-1 hairless (hr) mice with Ptchþ/ C57Bl/6 mice to generate a Ptchþ/ hr/ mouse [34]. They found enhanced tumor incidence, volume, and number per mouse in the hairless mice versus the haired. This was accompanied by an elevated inflammatory response and expression of mesenchymerelated proteins in the hairless mice. The relative expression levels of Hh signaling genes were not significantly different between groups, suggesting a similar pathogenesis of BCC tumor development [34]. It is of interest to note that the dose, duration, and frequency of UVB irradiation for haired and hairless mice was different; BCCs were induced with a higher and more frequent UVB dose (240 mJ/cm2) three times per week for 44 weeks in haired mice compared with hairless mice that were irradiated with 180 mJ/cm2 twice per week for 32 weeks [34]. This study additionally highlighted the difference in tumor susceptibly between males and females, wherein male mice developed more tumors than females which is similar to what is found in humans [35]. Another important factor that can influence BCC growth is the phase of the hair cycle. One study reported that BCC growth occurs preferentially during the anagen phase, perhaps due to elevated levels of Hh signaling [3]. These factors may be important in humans as well, considering that humans have relatively low level of hair in comparison to mice. BCC Chemoprevention: Pre-Clinical Studies Hh activation is crucial to BCC development; therefore, several Hh inhibitors have been investigated against BCC in order to reduce tumor progression. Systemic Hh signaling inhibition results in serious adverse effects such as muscle spasms, loss of taste, hair loss, fatigue, nausea, and hyponatremia [3]. One way to avoid some of these effects would be to use inhibitors topically to limit systemic exposure, although these topical applications may prove to be less efficacious [31]. One such Hh inhibitor, Cur61414, was shown to be effective in mice but when topically applied to human BCC, it could not clear the tumors and did not reduce Gli1 levels [31]. Another problematic issue with Hh inhibitors is the tumors’ ability to develop resistance over time. GDC0449, a Smo inhibitor, was used to treat a medulloblastoma patient, but slowly the tumor relapsed [36]. Upon investigation, it was discovered that an amino acid substitution in Smo was present which had no effect on the Hh signaling but disrupted the ability of GDC-0449 to bind to Smo [3,36]. Alternative Smo inhibitors are available that target wild type and the mutated portion of Smo, and these could be used to combat the drug resistance [37]. In addition to mutations found in Smo, other mechanisms of drug resistance involve amplification of downstream genes
CHEMOPREVENTION OF BCC
like gli2 and cyclin D1. Therefore, alternative therapeutic and chemoprevention strategies should be investigated to identify better options. In regard to studying the chemoprevention of BCC in pre-clinical models, the Ptch1þ/ mice present an excellent model mimicking the human disease condition. These mice develop similar tumors, have similar abnormalities in Hh activation, and the disease condition is developed in the same tissue (skin), and is induced through the same environmental insults (UV or IR) as in humans [38]. When Ptch1þ/ mice are irradiated with IR, they develop BCC and trichoepitheliomas, and when irradiated with UVB, they develop macroscopic tumors including BCC, SCC, and fibrosarcomas [38]. In Table 1, the BCC chemoprevention pre-clinical studies are summarized. There have been a relatively large number of studies targeting the BCC-induced inflammatory response including the use of non-steroidal anti-inflammatory drugs (NSAIDs) and sulfasalazine (SSZ). NSAIDs inhibit the cyclooxygenase (COX) enzymes (COX-1 and COX-2) responsible for generating prostaglandin mediated inflammation [38]. In one study, Ptch1þ/ mice were fed celecoxib (a specific COX-2 inhibitor) starting at 6 weeks of age, and then BCC was induced through a single dose of IR at 8 weeks of age. Celecoxib effect was not dose-dependent and when combining all doses used, it reduced microscopic BCC burden by 35% [38]. Additionally, the study investigated two other oral NSAIDs that were able to reduce microscopic BCC burden by 50 and 60%, sulindac and MFtricyclic, respectively. SSZ is a nuclear factor of kappa B (NF-kB) inhibitor and significantly decreased tumor number and volume in Ptch1þ/ hr/ mice [34]. Additionally, SSZ treatment decreased the level of Hh signaling components in BCC tumors and surrounding skin. The level of inflammation was also decreased by SSZ treatment, evidenced through decreased expression of p-p65, p-IKK a/b, and COX-2 [34]. There is an extensive amount of research on the chemopreventive effects of tea in SCC models showing promising effects in terms of inhibiting tumor promotion, malignant progression, and targeting various stages of multistep skin carcinogenesis [9]. However, a study conducted in a BCC mouse model testing green and black tea showed no efficacy, and even some higher number of tumors formed in female mice following green and black tea administration [39]. This study clearly highlights the difference in the pathogenesis of BCC and SCC, and also suggests that chemopreventive agents that are effective against SCC might not necessarily have similar efficacy against BCC. In mouse skin carcinogenesis, retinoids target the b-Raf/MAP kinase ERK kinase (MEK)/extracellular regulated MAP kinase (ERK) signaling pathway [4]. Tazarotene, an acetylenic retinoid, has been studied previously against SCC. It is currently approved in the United States for use in conditions such as topical Molecular Carcinogenesis
691
acne, psoriasis, and photoaging. One study showed the anti-BCC efficacy of tazarotene against BCCs induced by both UVB and IR [40]. Topically applied tazarotene inhibited the formation of microscopic BCC and reduced BCC size compared to vehicle. Tazarotene exerts its protective effects through inhibiting the phosphoinositide 3-kinase (PI3K)/ AKT/mammalian target of rapamycin (mTOR) pathway [32]. The importance of this pathway in BCC was evidenced through the use of pathway inhibitors that reduced the proliferation and tumorigenesis of BCC in vitro and in vivo [32]. Another target of interest in BCC is to investigate the DNA repair capacity given that DNA damage could lead to mutations in key molecules. Topical thymidine application as a potential chemopreventive has shown some promise [21]. A study investigated topically applied thymidine dinucleotide (pTT) in UVB irradiated Ptch1þ/ mice. pTT decreased the level of microscopic BCC, CPDs, 8-oxo-20 -dexoyguanosine (8oxo-dG), Ki-67 positive cells, p53 patches, and COX-2 [21]. Additionally, pTT increased the level of apoptosis in UVB-irradiated mice. Overall, this study showed pTT to be a promising chemopreventive drug that could target multiple pathways involved in BCC development. Ornithine decarboxylase (ODC) is another potential molecular target for chemoprevention of BCC. Overexpression of ODC in Ptchþ/ mice resulted in increased UVB-induced tumorigenesis, and pharmacologic inhibition of ODC with alphadifluoromethylornithine (DFMO) reduced tumor number per mouse as well as microscopic BCC area [41]. This study suggested utilizing DFMO as a potential chemopreventive agent against BCC. Vitamin D3 is another agent that has been found to reduce cell proliferation in BCC cells in culture and mice [42]. Ptchþ/ K14-Cre-ER p53fl/fl mice were treated primarily with tamoxifen to delete p53 expression, which accelerates murine BCC carcinogenesis. Mice were then exposed to IR at 8 weeks of age and BCCs formed by 5 to 6 months of age. After BCCs were visible, mice were treated with vitamin D3 for 30 days and tissue changes were visualized through immunohistochemistry. Overall, vitamin D3 treatment decreased the cell proliferation and also reduced the Gli1 mRNA levels [42]. Additionally, a study on calcitriol, the biologically active form of vitamin D3 has shown anti-BCC tumor growth effects in a conditional Ptch mutant mouse model [43]. These are promising results, and demand more studies focusing on the long-term chemopreventive effects of treatment with vitamin D3 and its active form calcitriol against BCC tumors. Silibinin, an active ingredient from the milk thistle plant seed extract, has shown extensive anti-cancer efficacy against various cancer types in animal models [44–51]. Additionally, there is extensive research on silibinin’s anti-cancer effects against SCC in the UVB-induced skin carcinogenesis model,
Molecular Carcinogenesis
2004 [41] 2011 [42] 2011 [43] 2014 [58]
2011 [31] 2010 [38] 2014 [34] 2001 [39] 2004 [40] 2008 [21]
Year (Ref)
Mitogenic signaling, NF-kB p50, c-Fos
VDR, Hh signaling
VDR, Hh signaling
Inhibitor of ODC
DNA repair
PI3K/AKT/mTOR pathway
Pleiotropic
Ptch1þ/ K14-CreER p53 fl/fl mice Ptchflox/flox ERT2þ/ mice Nude mice
Ptchþ/
Ptchþ/
Ptchþ/
Ptchþ/
Ptchþ/ hr/
NF-kB inhibitor
Smo inhibitor COX inhibitors
Mouse model Ptchþ/ K14-CreER2 p53fl/fl mice Ptchþ/ mice
Target
Injection of BCC cells
Tamoxifen
Single dose of IR
UVB 3 times per week
Single dose of IR or UVB 3 times per week UVB 3 times per week
UVB 3 times per week
Diet
Single dose of IR or UVB 3 times per week UVB 2 times per week
Decreased average tumor number and volume per mouse Ineffective
Inhibition of skin Hh signaling and BCC area decreased Reduced microscopic BCC burden
Results
Reduced formation of microscopic BCC Topical Decreased microscopic BCC area and number and level of CPDs, 8-oxo-dG, and Ki-67 were lowered Drinking Reduced number and area of water microscopic BCCs Topical Decreased cell proliferation and Gli1 expression in BCC tumors IP injection Inhibited cell proliferation and showed some decrease in BCC tumor area Oral Reduced tumor volume and inhibition gavage of proliferation as well as NF-kB and AP-1 activation
Drinking water Drinking water Topical
Topical
Treatment type
IR plus tamoxifen injection
Induction of BCC
BCC, basal cell carcinoma; Smo, smoothened; ptch, patch; IR, ionizing radiation; UVB, ultraviolet B; DFMO, difluoromethylornithine; ODC, ornithine decarboxylase; VDR, vitamin D receptor; Hh, hedgehog.
Silibinin
Calcitriol
Vitamin D3
DFMO
Thymidine dinucleotide
Tazarotene
Green and black tea
Celecoxib, Sulindac, MF-tricyclic Sulfasalazine
Cur61414
Agent(s)
Table 1. Summary of BCC Chemoprevention Pre-Clinical Studies
692 TILLEY ET AL.
CHEMOPREVENTION OF BCC
both dietary intake of silibinin as well as topically applied [45,52–57]. Recently, a study from our laboratory has shown that silibinin can inhibit BCC cell growth and clonogenicity in a dose-dependent manner [58]. Additionally, silibinin was shown to induce apoptosis and suggested to work via inhibiting multiple mitogenic signaling pathways (EGFR, ERK, AKT, NF-kB, etc.). In an ectopic allograft model, oral administration of silibinin strongly inhibited BCC cells tumor growth. These initial studies have provided evidence that silibinin can potentially be used as a chemopreventive agent against BCC but additional studies still need to be conducted. BCC Chemoprevention: Clinical Studies There has not been great success in chemoprevention trials of BCC in humans, the various studies summarized herein are depicted in Table 2. In the past, several trials have aimed to analyze the incidences of both SCC and BCC, which can be challenging in itself since the two types of NMSC are so different. One study conducted by the Skin Cancer Prevention study group administered 50 mg dose of b-carotene daily for 5 years to patients with a recent NMSC, and resulted in no significant effect on the incidence of new NMSC [5,59]. Another group looked at the Nutritional Prevention of Cancer study group with 1,312 patients with a history of NMSC. These patients were treated with 200 mg of selenium or placebo for about 4 years with a 6 year follow-up, wherein selenium showed no effect on incidence of BCC or SCC [5,60]. The SW Skin Cancer Prevention study group enrolled 2,297 patients with a history of greater than 10 AK and less than 2 SCC or BCC. They treated the patients with retinol for 5 years and conducted a 3.8 year follow-up but did not observe any effect on BCC incidence [5,61]. Another study conducted in Australia using 30 mg of daily b-carotene and sunscreen showed no effect on the incidence of BCC, though sunscreen use appeared to reduce number of SCC after 4.5 years follow-up [5,62]. These large group studies have thus far shown no effective chemopreventive option for BCC. Studies with retinoids in clinical models have shown a chemopreventive effect of oral retinoids for SCC, but there is not enough evidence for their use to decrease the risk of BCC [1,4,6,63,64]. Retinoids are considered to be natural and synthetic derivatives of vitamin A. They regulate important functions like signaling for growth arrest, differentiation, and apoptosis; additionally, they modify expression of growth factors, signal transduction, extracellular adhesion molecules, and proto-oncogenes [4]. These compounds act through retinoic acid receptors and retinoid x receptors [1,6]. Additionally, the agents must be continued indefinitely to maintain their protective benefits [4]. The side effects and efficacy of oral retinoids are dose-dependent. Side effects include mucocutaneous dryness and irritation, hair loss, hypertriglyceridemia, hypercholesterolemia, liver toxicity, increased intracranial pressure, and Molecular Carcinogenesis
693
skeletal changes [6]. The use of retinoids for chemoprevention is to decrease the number of NMSC to a manageable level. There are two oral retinoids available in the United States, isotretinoin, and acitretin. Acitretin is the primary choice for chemoprevention. Topical retinoids have a response rate of about 50%, therefore, the routine use of topical retinoids for chemoprevention is not recommended. In the Veterans Affairs Topical Tretinoin Chemoprevention Trial (VATTC), patients were administered a high dose of topical tretinoin for 1.5–5.5 years [64]. Tretinoin was ineffective at reducing the risk of new cases of NMSC. Topical tazarotene was tested in a phase II, randomized, double blind, vehicle control study in 34 basal cell nevus syndrome (BCNS) patients, but did not show significant anti-BCC effects [63]. Overall, systemic retinoids appear to have a modest effect on existing tumors and can suppress new tumor development, while topically applied retinoids have not been as effective [6]. Photodynamic therapy (PDT) has been found to reduce the number of new cases of AK but the chemopreventive effect of PDT on NMSC is inconclusive [1]. Cyclic 5-aminolevulinic acid (ALA)-PDT for the prevention of NMSC in high risk organ transplant recipients (OTRs) showed that treatment at 4 to 8 week intervals for 2 years caused significantly fewer incidences of SCC than in the year before treatment [4]. Topical application of methyl aminolevulinate (MAL) and ALA result in the accumulation of protoporphyin IX preferentially in damaged skin lesions. Subsequent irradiation of treated tissue with PDT resulted in high intra-lesional concentrations of reactive oxygen species (ROS) and selective tissue destruction [4]. A case report on a patient with BCNS found MAL-PDT to be an effective chemopreventive agent against new BCC development, though this is a single patient study and needs to be validated in more patients before any conclusions can be drawn from this study [65]. Another agent that could be possibly useful against BCC is 5-fluorouracil (5-FU). 5-FU is approved for treatment of AKs and the 5% cream is approved for superficial BCCs. There is evidence for effectiveness of 5-FU in reducing AKs but it has not been studied as a chemopreventive agent against NMSC. NSAIDs show conflicting results in literature for the chemoprevention of BCC [38,66–69]. A study evaluating data from a clinical practice research database compared prior exposure to systemic NSAIDs between patients with incident BCC and SCC along with patients without NMSC [69]. They found no evidence for reduced risk of BCC in association with systemic NSAIDs use, though long-term use of ibuprofen or aspirin was associated with a small BCC reduction. One randomized, placebo-controlled, double blind trial evaluated the efficacy of celecoxib as a chemopreventive agent for AK in patients with multiple AK; they found after 9 months of treatment, celecoxib
Molecular Carcinogenesis
2014 [69]
2010 [67]
2010 [38]
2010 [70] 2001 [73]
NSAIDs
Celecoxib
Celecoxib
DFMO 30 1,051
Population at very high risk for NMSC
291
60
240
>60,000
1
34
XP
History of prior NMSC
BCNS
10–40 AKs
BCC and SCC diagnosed between 1995 and 2013
BCNS
BCNS
1131
1621
2297
1312
1805
Number of participants Oral administration/5 years
Treatment/duration
Various chemopreventive treatments including MAL-PDT for a few sessions Identified patients with diagnosis of NMSC and assessed prescriptions for systemic NSAIDs Oral administration/9 months
Topically applied/follow-up of 3 years
Topically applied/1.5 to 5.5 years
Oral administration/4.5 years
NSAIDs ineffective at reducing BCC, but reduced BCC risk in regular users of aspirin and ibuprofen Fewer NMSC in the celecoxib arm than in the placebo arm
Effective at reducing new BCCs in single patient
Ineffective
Ineffective
Ineffective
Ineffective
Ineffective
Ineffective
Results
Cohort study design
Analyzed patients taking ACE inhibitors or ARBs/3.4 years
Reduced risks of BCC and SCC in patients taking ACE or ARBs
Phase II-III, double-blind, placebo-controlled, randomized trial Phase II, double-blind, Oral administration/24 months Reduced BCC burden only in patients placebo-controlled, randomized with
