FormalPara Key Points

Sonidegib and vismodegib are Hedgehog pathway inhibitor medications approved to treat patients with advanced basal cell carcinoma that cannot be adequately treated with surgery or radiation; while these agents are often useful for complicated cases, they are also associated with undesirable side effects that may cause patients to stop treatment before the medication has had optimal effect.

The most common side effects seen with both sonidegib and vismodegib treatment include muscle spasms, hair loss, abnormal taste perception, nausea, severe tiredness, diarrhea, weight loss, and poor appetite.

Strategies to reduce the side effects associated with vismodegib and sonidegib include temporary treatment interruptions, dose reductions, and symptom-specific therapies; healthcare practitioners should also educate patients about the possible side effects of these medications so that they are less likely to stop treatment prematurely.

1 Introduction

Basal cell carcinoma (BCC) is the most common cancer in fair-skinned adults worldwide, representing approximately 70–80% of all non-melanoma skin cancer (NMSC) cases [1, 2]. BCC is typically observed in patients over 50 years of age, with an estimated median patient age of 66 years in the USA [3]. Risk factors include ultraviolet radiation exposure, increasing age, immunosuppression, male sex, White race, and history of prior NMSC [1, 3, 4].

BCC lesions are usually slow growing and commonly found on the face, trunk, and extremities [4, 5]. The mortality rate for BCC is low, but prominent location, high recurrence rate, and potential for local tissue invasion can cause significant morbidity and emotional distress in affected patients [4,5,6]. Advanced BCC comprises locally advanced BCC (laBCC), defined as primary tumors that invade surrounding tissues including local lymph nodes, and metastatic BCC (mBCC), lesions that spread from the primary tumor to distant sites, most commonly the lymph nodes, lungs, and bone [7]. Advanced BCC tends to occur in a small subgroup of patients who have often forgone treatment for years or experienced treatment failures or recurrences [8]. While surgery is the first-line treatment for localized BCC, treating advanced BCC can be challenging, as traditional therapies such as surgery and radiation may not be viable options [5, 7, 9].

Hedgehog pathway inhibitors (HHIs) target mutations in the Hedgehog pathway that are associated with BCC pathogenesis and may be used in patients with advanced BCC not amenable to traditional therapeutic options [10, 11]. Currently, there are two HHIs approved for the treatment of advanced BCC—sonidegib 200 mg daily (ODOMZO®; Sun Pharmaceutical Industries, Inc.) and vismodegib 150 mg daily (ERIVEDGE®; Genentech, Inc.). These agents are both approved by the US Food and Drug Administration (FDA), the European Medicines Agency (EMA), the Australian Therapeutic Goods Administration (TGA), and Swissmedic for the treatment of adult patients with laBCC that has recurred following surgery or radiation therapy, or those who are not candidates for surgery or radiation therapy; both agents are also approved by the TGA and Swissmedic for the treatment of mBCC, and vismodegib is approved by the FDA and EMA for the treatment of adults with mBCC [12,13,14,15,16,17,18,19,20]. Vismodegib was approved by the FDA for BCC treatment first in 2012, and sonidegib received FDA approval for this indication in 2015 [21, 22].

While there has been demonstrated success with both of these HHIs in the treatment of advanced BCC, the use of these agents is sometimes limited by treatment-related adverse events (AEs), necessitating the establishment of AE management strategies to enable continuous treatment and maximize therapeutic benefit [10, 12]. This review compares the safety and tolerability of the approved agents sonidegib and vismodegib in the treatment of BCC and discusses AE management strategies that may improve HHI tolerability in affected patients.

2 Basal Cell Carcinoma Diagnosis and Classification

The diagnosis of BCC involves visual and manual examination, dermatoscopy to detect characteristic features, and biopsy of the lesion and any involved lymph nodes or distant metastases for histopathologic confirmation [4, 23]. BCC staging entails assessment of the primary tumor site and determination of local invasion and/or distant metastasis and may utilize multiple different staging systems [23, 24]. The American Joint Committee on Cancer tumor, node, metastasis (TNM) staging system is used mostly for patients with advanced BCC, rather than localized lesions [8, 23, 25]. The European Association of Dermato-Oncology utilizes a four-stage system developed for improved clinical classification of BCC based on difficulty of treatment and recurrence risk [26]. The National Comprehensive Cancer Network proposes different primary and additional treatment recommendations for BCC based on risk status at diagnosis (low risk, high risk, locally advanced disease, and regional or distant metastatic disease) [24]. Ocular and periocular BCCs are classified using the Union for International Cancer Control TNM 8th edition staging, which describes size/extent of invasion of the primary tumor and extent of regional or distant metastases [27].

The goal of BCC treatment is complete removal of the tumor while maximally preserving function and cosmesis at the lesion site [4]. Available treatment options for localized lesions include traditional excisional surgery, Mohs micrographic surgery, curettage and electrodesiccation, cryosurgery, topical therapy, laser, photodynamic therapy, and radiation therapy [25]. Surgical resection of most primary BCCs on the trunk and extremities is effective, given that tissue conservation is not a priority [4]. For advanced BCCs that become locally invasive or metastatic, surgical resection may not be feasible given the tumor size or number, or proximity to vital or functionally important structures [4, 24]. For example, surgical excision of laBCCs on the head and neck could cause disfiguration and loss of function [12]. Thus, patients with advanced BCC have a need for non-invasive alternative treatment options [9].

3 Hedgehog Pathway Inhibitors in the Treatment of Basal Cell Carcinoma

The Hedgehog signaling pathway regulates cell growth and differentiation during development, cell proliferation, and tissue repair [10]. Pathway signaling is initiated via binding of a Hedgehog ligand (Sonic, Indian, or Desert Hedgehog) to the Patched-1 receptor, thereby relieving its inhibition of the Smoothened receptor (Fig. 1) [10]. Smoothened receptor signaling activates the glioma-associated oncogene homolog 1 (Gli1) transcription factor, leading to the induction of Hedgehog target genes, including GLI1, Patched-1, and other genes regulating the cell cycle or angiogenesis [28]. The development of BCC is associated with mutations causing upregulation of the usually inactive Hedgehog pathway, allowing aberrant inhibition of suppressor genes or enhanced transcription of activating genes [10, 28, 29]. Most (~ 80%) sporadic BCCs have loss-of-function mutations in Patched-1 (suppressor), and others have gain-of-function mutations in Smoothened (activator) leading to dysregulation of canonical pathway signaling and uncontrolled proliferation of epidermal basal cells [10, 30].

Fig. 1.
figure 1

Hedgehog pathway. a Normal Hedgehog pathway signaling; b effect of Hedgehog pathway inhibitors. GLI1 glioma-associated oncogene homolog 1, HH Hedgehog, PTCH-1 Patched-1, Smo Smoothened

Full size image

HHIs were developed from the naturally occurring, teratogenic HHI cyclopamine and inhibit Smoothened, thereby blocking aberrant downstream Hedgehog signaling and BCC tumorigenesis [31]. The poor tolerability of cyclopamine in animal models led to the development of HHI derivatives with improved specificity, potency, and safety profiles [31,32,33].

Sonidegib and vismodegib are the only HHIs approved for advanced BCC to date, but several other HHIs have been investigated for potential treatment of patients with BCC. These include patidegib (saridegib, IPI-926), taladegib (LY2940680, ENV-101), and itraconazole. Patidegib is a semisynthetic HHI available as a gel formulation, which received orphan drug designation for the treatment of nevoid BCC syndrome (Gorlin syndrome) [34, 35]. A Phase II trial in patients with non-Gorlin high-frequency BCC was terminated early due to low blinded event rate, and there are currently no active BCC trials [36]. Phase I studies of the selective oral agent taladegib in patients with advanced solid tumors including BCC are complete, and a Phase II study in patients with refractory advanced solid tumors and Patched-1 loss-of-function mutations is currently recruiting [37,38,39]. Both oral and topical formulations of itraconazole, an azole antifungal agent, have been investigated in patients with BCC in Phase II studies and case reports [40,41,42,43].

Due to the challenges associated with treating advanced BCC, neoadjuvant use of HHIs, including sonidegib and vismodegib, is also being investigated, though neither of these agents are approved for this indication [44, 45]. Neoadjuvant HHI therapy aims to reduce the size of advanced tumors prior to curative surgical excision, radiation therapy, or topical treatment [44, 46]. The literature regarding neoadjuvant use of sonidegib in BCC is currently limited to observational studies and case reports with promising results [9, 47,48,49], and a clinical trial evaluating neoadjuvant sonidegib followed by surgery or imiquimod for BCC is recruiting as of March 2024 [46]. Multiple clinical trials, observational studies, and case reports have also demonstrated the efficacy of vismodegib as a neoadjuvant therapeutic in BCC [47, 49,50,51,52,53,54,55,56,57,58,59,60]. For example, in the Phase II, open-label VISMONEO study enrolling 55 patients with laBCC, vismodegib treatment led to surgical downstaging in 44 (80.0%) patients, with a histologically confirmed complete response to vismodegib observed in 27 (49%) patients. Subsequently, 23 patients underwent closing surgery, and 1 patient underwent radiotherapy. In the following sections, we compare the pharmacokinetic (PK), safety, and tolerability profiles of sonidegib and vismodegib for their currently approved indications in BCC.

4 Pharmacokinetic Profiles of Sonidegib and Vismodegib

Sonidegib and vismodegib are both highly bound to plasma proteins (> 97% and 97%, respectively, in vitro) and have long elimination half-lives (28 days and 4–12 days, respectively) [13,14,15, 20]. However, there are also several notable differences in the PK profiles of these agents [11, 61]. Vismodegib exhibits a non-linear PK profile with concentration-dependent changes and has a mean absolute bioavailability of 31.8% after a single dose and ~7% after continuous once-daily dosing [62, 63]. Sonidegib exhibits a dose-proportional PK profile within the dose range of 100–400 mg, which includes the approved 200 mg dose, and has a bioavailability of < 10% [13, 14, 64, 65]. The absorption of sonidegib, but not vismodegib, is affected by food, and drug exposure increases upon administration with a high-fat meal [13, 20].

Steady-state drug levels are reached in 4 months with sonidegib administration and in 7–21 days with vismodegib administration [13, 14, 63]. Vismodegib has a low volume of distribution (16.4–26.8 L), indicating confinement to the plasma, whereas sonidegib has a high volume of distribution (> 9000 L), indicating extensive tissue penetration [12, 62, 66]. As such, sonidegib concentration is six times higher in the skin than the plasma [13]. Whether these differences in the sonidegib and vismodegib PK profiles translate into clinical differences between the two agents is unclear [61].

5 Safety and Tolerability

5.1 Common Adverse Events

In the pivotal Phase II BOLT study (NCT01327053), sonidegib demonstrated a manageable safety profile that was more favorable for the approved 200 mg dose than the 800 mg dose [67, 68]. Treatment-emergent AEs were reported in almost all patients receiving the 200 mg dose as of the primary analysis at 6 months (95%) and the final analysis at 42 months (98%; Table 1) [67, 68]. The most common (≥ 20%) AEs of any grade observed with sonidegib 200 mg at the 42-month analysis were muscle spasm, alopecia, dysgeusia, nausea, fatigue, diarrhea, decreased weight, creatine phosphokinase (CK) increase, and decreased appetite (Table 2) [67]. The long-term safety profile of sonidegib in BOLT was comparable to that observed at the primary 6-month analysis, with no new safety concerns reported after 42 months of follow-up [67].

Table 1 Overview of Hedgehog pathway inhibitors approved for the treatment of BCC
Full size table
Table 2 Incidence and time to onset of the most common AEs with vismodegib and sonidegib treatment
Full size table

In the pivotal Phase II ERIVANCE trial (NCT00833417), vismodegib 150 mg had an acceptable safety profile [69]. However, treatment-emergent AEs were reported in 100% of patients as of the primary analysis at 9 months (Table 1) [69]. The most common (≥ 20%) AEs of any grade were similar to those observed with sonidegib 200 mg and included muscle spasms, alopecia, dysgeusia, weight decrease, fatigue, nausea, decreased appetite, and diarrhea (Table 2) [69, 70]. These AEs, along with constipation, arthralgia, vomiting, dyspepsia, pruritis, rash, ageusia, and pain (general and in extremities), were considered the most common (≥ 10%) AEs of any grade in a pooled analysis of safety data from four Phase I or II vismodegib clinical trials, including ERIVANCE [15, 20]. The long-term vismodegib safety profile remained consistent with that of the primary analysis; however, the incidence of the most common AEs generally increased with longer exposure durations (≥ 12 months) [70].

Treatment-emergent AEs were a common reason for treatment discontinuation in BOLT and ERIVANCE [68, 69]. In the BOLT final analysis, the AEs most frequently leading to sonidegib treatment discontinuation were muscle spasms (5%; 4/79); asthenia, dysgeusia, and nausea (each 4%; 3/79); and fatigue, weight loss, and decreased appetite (each 3%; 2/79) [67]. The most common AEs leading to vismodegib treatment discontinuation as of the ERIVANCE final analysis were muscle spasms (4.8%; 5/104) and weight decrease and dysgeusia (each 1.9%; 2/104) [70]. These AEs typically appeared within the first few weeks of treatment with both sonidegib and vismodegib (Table 2) [12, 61].

When examining these pivotal trial data in the absence of head-to-head comparisons, the safety profiles for sonidegib and vismodegib appear similar [61]. Sonidegib may be associated with less severe and less frequent treatment-emergent AEs overall compared with vismodegib [61, 72, 73]. Patients on sonidegib may have lower incidences of muscle spasm, alopecia, and dysgeusia versus vismodegib, but higher incidences of nausea, vomiting, fatigue, diarrhea, decreased appetite, and myalgias [11, 71, 72, 74, 75]. Additionally, most AEs (except fatigue and weight decrease) were found to have later onset with sonidegib than vismodegib, though these differences were not statistically significant [11]. The shortest median times to AE onset with vismodegib treatment in ERIVANCE were for dysgeusia (1.48 months) and muscle spasm (1.89 months); for sonidegib in BOLT, the shortest median times to AE onset were for fatigue (1.08 months) and muscle spasm (2.07 months) [11, 61, 76]. In BOLT, most AEs associated with sonidegib were reported to be manageable or reversible with dose interruptions or reductions, though the times to resolution of common AEs are not provided [67]. An exploratory analysis of 266 patients who completed a 12-month follow-up in the post-approval, open-label STEVIE trial (NCT01367665) of vismodegib 150 mg in laBCC and mBCC reported that 97% of patients had ≥ 1 ongoing AE at time of treatment discontinuation [77]. Most occurrences of muscle spasm resolved within 1–3 months, most incidences of ageusia, dysgeusia, and alopecia resolved within 6 months, and most instances of decreased weight resolved within 12 months after stopping vismodegib [71, 77].

Given the similarities in overall AE profiles observed with vismodegib and sonidegib, including the high incidence of muscle spasms, alopecia, dysgeusia, nausea, decreased appetite, and fatigue, these common AEs may be considered class effects with HHIs [7, 12, 61]. Early patient education by healthcare practitioners about potential AEs before starting HHI treatment and effective AE management strategies during treatment are needed to prevent treatment discontinuation and optimize clinical outcomes [61, 78].

5.2 Adverse Events of Interest

Though not considered common in sonidegib or vismodegib clinical trials, other AEs of interest with HHI treatment should also be addressed. For example, amenorrhea has often been observed in females receiving HHIs. The Hedgehog pathway is vital for ovarian steroidogenesis; thus, amenorrhea is thought to occur with HHI therapy due to inhibition of ovarian function causing abnormal follicle-stimulating hormone-dependent signaling [79, 80]. Amenorrhea may occur in up to 50% of females of childbearing potential taking vismodegib but has not been commonly reported with sonidegib [81]. At the ERIVANCE 12-month update, amenorrhea was reported in two of the six females of childbearing potential [76, 82]. Additionally, in STEVIE, 8 of the 29 enrolled females of childbearing potential experienced irregular menses or amenorrhea [83]. In an expanded access trial (NCT01160250) of vismodegib in patients with advanced BCC, four of eight females of childbearing potential reported amenorrhea or irregular menses [84]. A systematic review and pooled analysis of HHI studies by Jacobsen et al. (2016) found that the incidence of amenorrhea ranged from 27.6 to 100% in patients with BCC receiving vismodegib, with a weighted average of 32.9% [85]. Multiple case studies have also described amenorrhea as an AE associated with vismodegib treatment [79, 80, 86,87,88]. In the BOLT primary analysis, Grade 3 amenorrhea was reported in a single patient receiving sonidegib 200 mg; the total number of females of childbearing potential in BOLT is not reported [68]. Amenorrhea secondary to HHI treatment is thought to be reversible, as affected patients have experienced a return to normal menses in the weeks following treatment cessation [79,80,81, 87, 89]. Additionally, while pregnancy during HHI therapy is contraindicated, the literature regarding fertility following HHI therapy is limited [13,14,15, 20]. Summaries of product characteristics for both vismodegib and sonidegib state that fertility may be compromised by treatment, as based on data from animal studies [14, 15, 90].

Hepatotoxicity has been reported in postmarketing surveillance as a severe AE in patients taking vismodegib, though this was not commonly reported in clinical trials [20, 91, 92]. The proposed mechanism of liver damage with vismodegib is via drug–drug interactions, direct hepatocellular damage, or impaired hepatocyte regeneration [81]. Two analyses reviewed vismodegib data from the FDA Adverse Events Reporting System (FAERS), a publicly accessible, voluntary database of self-reported AE and medication error incidents reported to the FDA, to investigate the real-world incidence of hepatotoxicity in patients receiving this agent [91, 92]. Ventarola and Silverstein reviewed FAERS data from 2012, the year that vismodegib was FDA-approved, and found that liver toxicity represented 15 (23.1%) of the 65 individual cases of vismodegib-associated reactions [92]. Vismodegib was noted as the agent responsible for hepatotoxicity in 4 of these cases (26.67%); the remaining 11 cases cited other agents as also potentially responsible [92]. Edwards et al. [91] also searched FAERS data (from 1 January 2009–31 December 2015) and found 94 cases reported as having ≥ 1 AE of liver dysfunction with vismodegib; 35 of these included terms suggesting severe hepatotoxicity [91]. The analysis also identified an increasing incidence of hepatic AEs reported to FAERS over the study period [91]. However, it is important to consider that these FAERS reports do not generally provide detail regarding patient factors potentially contributing to the AE, such as existing liver diseases or abnormalities, alcoholism, obesity, or use of concomitant medications [91]. Multiple case reports have described hepatic abnormalities in patients taking vismodegib, with the first being published in 2011 [93,94,95,96,97,98,99]. The EMA classifies vismodegib as potentially hepatotoxic, and drug-induced liver injury is listed as an AE noted during post-approval use of vismodegib in the US Prescribing Information [10, 20]. Thus far, the use of sonidegib has not been associated with hepatotoxicity; in BOLT, one Grade 3 alanine aminotransferase elevation and Grade 3 aspartate aminotransferase elevation were each noted in one patient receiving sonidegib 200 mg [68]. However, the discrepancy in reports of hepatotoxicity with vismodegib versus sonidegib may reflect the real-world use of these agents, as vismodegib was FDA-approved for the treatment of advanced BCC 3 years earlier than sonidegib.

Patients receiving HHIs for the treatment of BCC may also be at increased risk for the development of squamous cell carcinoma (SCC). In ERIVANCE, SCC was reported as an AE in 12 (11.5%) patients receiving vismodegib 150 mg, and in BOLT, SCC as a secondary malignancy was noted in 3 (3.8%) patients receiving sonidegib 200 mg [68, 70]. In STEVIE, a total of 51 (4%) patients with laBCC or mBCC receiving vismodegib 150 mg daily reported cutaneous SCC. Most of these patients were > 75 years of age and had lesions located in sun-exposed areas, and 18 of the 51 (35%) patients had a history of cutaneous SCC [77]. A case–control study by Mohan et al. (2016) in patients exposed versus not exposed to vismodegib reported a significant risk of SCC with vismodegib therapy [100]. Jacobsen et al. (2016) found that onset of new SCC was reported in 0.8–20.0% of patients with BCC on vismodegib, with a weighted average of 1.3% [85]. Additionally, several case reports have described patients with new development of SCC during or after vismodegib therapy for BCC [86, 101,102,103,104,105,106]. However, a retrospective cohort study of patients receiving vismodegib in Phase I and II clinical studies compared with patients receiving surgical BCC treatment did not find a significant difference in SCC incidence between the two groups [107]. In addition, a systematic review and meta-analysis found that a new diagnosis of SCC was made in 3.6% of patients on vismodegib, which was not statistically significant (P = 0.77), and estimates were smaller in larger studies [75]. Other than the BOLT primary analysis, no published studies or case reports describe an association between SCC and sonidegib treatment. The exact mechanism for the potential risk of SCC with HHI treatment is unknown [81, 100]. Inhibiting Smoothened may promote the growth of tumor cells that can circumvent the Hedgehog pathway via alternate mechanisms, such as activation of the RAS/MAPK pathway [81, 100]. However, the observed occurrence of SCC in patients with BCC may be confounded by the fact that BCCs and SCCs share a common risk factor of ultraviolet (UV) light exposure [75, 81, 100]. Additionally, as patients with BCC undergo more frequent skin examinations than the normal population, the discovery of SCC in these patients may represent a detection bias [75, 81, 108]. It should be noted that, in Mohan et al. (2016), patients with BCC were examined more frequently than control patients due to monitoring as part of clinical studies or for treatment-emergent AEs [100, 108].

5.3 Muscle Spasms and Creatine Phosphokinase Elevations

Muscle spasms usually occur early in treatment with HHIs, worsen with length of therapy, and are a common reason for patient discontinuation [11, 109, 110]. Two underlying mechanisms have been proposed to explain the association between HHIs and muscle spasms (Fig. 2). In the first mechanism, HHI-induced Smoothened inhibition activates noncanonical Hedgehog signaling pathways, which enhances calcium influx into myocytes and increases muscle contractility [10, 110, 111]. The other proposed mechanism involves myofibroblasts, which act as a scaffold for skeletal muscle cells [110]. Hedgehog pathway inhibition decreases the expression of alpha smooth muscle actin (α-SMA) and production of E-cadherin by myofibroblasts, leading to disruption of the actin cytoskeleton and to muscle contraction [110]. This hypothesis is particularly relevant to aggressive BCC subtypes, which have increased α-SMA expression correlating with tumor invasiveness [110].

Fig. 2
figure 2

Potential mechanisms for HHI-associated muscle spasms. HHI inhibition of Smo may lead to muscle contractions by two proposed mechanisms. In mechanism 1, noncanonical Hedgehog pathway activation leads to increased calcium influx and myocyte contraction [10, 110, 111]. In mechanism 2, decreased transcription of α-SMA and production of E-cadherin by myofibroblasts leads to dissociation of the myocyte–myofibroblast junctions that act as a scaffold for myocytes, thereby triggering myocyte contraction [110]. α-SMA alpha smooth muscle actin, HHI Hedgehog pathway inhibitor, Gli glioma-associated oncogene homolog 1, Smo Smoothened

Full size image

CK elevations may be associated with HHI-induced muscle spasms [12, 13, 113]. Increased CK, which is typically preceded by musculoskeletal pain and myalgia, was a common AE reported with sonidegib 200 mg treatment in BOLT, occurring in 29.1% and 30.4% of patients in the primary and final analyses, respectively (Table 2) [67, 68]. Median time to Grade > 2 CK elevation was approximately 2.58 months following treatment initiation, and Grade ≥ 2 events resolved to Grade ≤ 1 in a median of 12 days (range 8–14 days) [13, 61, 114]. Of note, CK levels were not monitored as part of the laboratory assessments in patients receiving vismodegib in ERIVANCE [89]. In STEVIE, an exploratory analysis found that 60 of the 432 patients who received vismodegib 150 mg and had ≥ 1 CK measurement available during the study treatment period had elevated CK values [77]. In the randomized, double-blind, Phase II MIKIE study (NCT01815840) assessing two long-term intermittent vismodegib 150 mg dosing regimens in patients with multiple BCCs, 9% (10/114) of patients in the first treatment group (three cycles of 12 weeks of vismodegib followed by 8 weeks of placebo, and then 12 final weeks of vismodegib) and 10% (11/113) of patients in the second treatment group (24 weeks of vismodegib followed by three cycles of 8 weeks of placebo and 8 weeks of vismodegib) experienced Grade 1–2 CK increase, and Grade 3 CK elevations occurred in 1% (n = 1) and 4% (n = 4) of patients, respectively [115]. Time to CK elevation with vismodegib treatment was not reported in STEVIE or MIKIE.

Increased CK is considered a class-related AE of HHIs [12]. Elevations in serum CK can be biomarkers of muscle damage following exercise and in patients with myocardial infarction, muscular dystrophy, cerebral diseases, or peripheral neuropathies [116, 117]. Rhabdomyolysis was reported by study investigators as an AE in one (1.3%) patient receiving sonidegib 200 mg at the 30-month BOLT analysis, but this case was not confirmed by an independent safety review and adjudication committee on muscle toxicity [68, 114]. No cases of rhabdomyolysis were reported in vismodegib clinical trials, but as previously stated, CK levels were not monitored in ERIVANCE [77]. In BOLT, 38.2% (13/34) of patients receiving sonidegib without CK increase experienced muscle spasms, compared with 53.1–75.0% of patients with Grade 1 (17/32) or Grade 2 (6/8) CK elevation [68]. In STEVIE, CK elevations were noted in 36.4% (44/432) of patients without muscle spasms and 33.9% (20/432) of patients with muscle spasms while receiving vismodegib [77]. Therefore, it is unclear whether the occurrence of increased CK is associated with muscle-related AEs in patients with advanced BCC receiving HHIs.

Though most HHI-associated AEs are low grade, their chronicity and aggregation may decrease patient quality of life and lead to treatment discontinuation [68, 69, 109]. Therefore, implementing AE mitigation strategies prior to and during treatment, as well as at AE onset, is important in managing the therapeutic course of HHIs [12].

6 Adverse Event Management Strategies

6.1 Patient Evaluation and Monitoring

Prior to treatment initiation, healthcare providers should obtain a thorough patient history and physical examination to elucidate comorbidities that may contribute to the likelihood of AEs with HHIs, such as pre-existing alopecia, myopathies, diabetes, gastrointestinal/liver disease, and sensory/motor neuropathies [89, 109]. Certain concomitant medications may decrease tolerability of HHIs or increase risk of AEs (e.g., myotoxic medications and statins) [89, 109]. Additionally, as AEs tend to occur early in the course of treatment, providing adequate patient education prior to treatment initiation can help manage expectations regarding AEs and maximize treatment adherence [12, 78, 109].

As of March 2023, serum CK monitoring and renal function tests are recommended in the prescribing information for both sonidegib and vismodegib [13, 20]. Serum CK levels should be measured at baseline and periodically during treatment or as clinically indicated with occurrence of musculoskeletal AEs [13, 20]. Depending on the severity of musculoskeletal AEs or CK elevation, dose interruption or discontinuation may be required [13, 20]. In addition, though not currently recommended in the product prescribing information, practitioners may consider monitoring for liver enzyme elevations during HHI treatment due to the occurrence of drug-induced liver injury observed postmarketing with vismodegib [20].

6.2 Strategies for Adverse Event Management

Management of common AEs can involve pharmacologic options, supportive care approaches, and existing treatment modification; these are detailed in Table 3. As muscle spasms, alopecia, weight loss, and dysgeusia are the most frequent AEs reported with HHI treatment and often lead to discontinuation, aiming to reduce the incidence of these AEs specifically may have the most significant impact in optimizing treatment adherence [11].

Table 3 Management strategies for class-effect AEs associated with Hedgehog pathway inhibitors
Full size table

6.3 Treatment Interruptions or Dose Reductions

Treatment interruptions (breaks in HHI therapy for multiple weeks at a time) are a widely used strategy to prevent discontinuation of HHIs due to associated AEs [12, 118, 119]. To date, no studies have evaluated the effects of prespecified dose interruptions with sonidegib treatment, and additional research is needed to determine whether treatment interruptions or dose reductions (alternate-day to alternate-week dosing) would impact sonidegib efficacy [12, 120,121,122]. However, the high tissue distribution and extended half-life of sonidegib may create opportunity for alternative dosing paradigms [12, 13, 61]. In BOLT, 68.4% and 16.5% of patients receiving the approved sonidegib 200 mg dose had a dose interruption or reduction, respectively [118]. Objective response rates were comparable for patients with laBCC receiving sonidegib 200 mg regardless of dose reduction/interruption status (50.0% for patients with ≥ 1 dose reduction/interruption and 57.4% for patients without any dose reductions/interruptions) [118]. Interruptions of 2–4 weeks in duration may therefore be recommended for management of CK elevations or Grade ≥ 3 muscle spasms that occur with sonidegib treatment (Table 4) [109, 118]. Multiple interruption cycles may be required, so discontinuing treatment is not recommended until at least two interruption cycles of 2–4 weeks each have been attempted [109]. In BOLT, dose reductions were only considered in the case of AE recurrence after sonidegib reinitiation following a treatment interruption; in this case, the dosage was reduced on second reinitiation of the study drug [12, 118]. A maximum of one dose reduction (to placebo) was permitted in patients receiving sonidegib 200 mg; if further reductions were necessary, the patient was discontinued from study treatment [12]. Dose reductions to one 200 mg capsule every other day as required to reduce AEs are specified within the sonidegib EMA product label [15, 78, 119].

Table 4 Recommended dose interruptions, dose modifications, and monitoring for suspected treatment-related CK increase with or without symptoms during sonidegib treatment [12, 118]. Table adapted with permission from Lewis [118] (http://creativecommons.org/licenses/by-nc/4.0/) and Lear [12] (https://creativecommons.org/licenses/by/3.0/)
Full size table

Alternative treatment dosing regimens for vismodegib have also been evaluated, and per the US product label, dose interruptions of up to 8 weeks in duration may be recommended for intolerable AEs until improvement or resolution [20]. Evidence suggests that, similar to sonidegib, treatment interruptions do not impact vismodegib efficacy [70, 115, 123]. In an ERIVANCE exploratory analysis, a comparable proportion of patients with laBCC having no missed doses (58.3%) and patients having up to 1/3 missed doses (63.3%) achieved objective treatment responses; in the mBCC cohort, objective response was noted in 60.0% of patients with no missed doses and in 43.5% of patients with up to 1/3 of doses missed [70]. In MIKIE, two intermittent dosing regimens, each lasting a total of 72 weeks with multiple 8-week treatment interruptions (48 weeks of vismodegib 150 mg and 24 weeks of placebo total), produced similar mean relative reductions in the number of BCC lesions from baseline to week 73 [115]. A post hoc analysis of ERIVANCE and STEVIE, using a tumor growth inhibition model describing longitudinal tumor size change over time, demonstrated that 8-week treatment interruptions had no impact on the proportion of patients predicted to have a treatment response with vismodegib [123]. Further, a post hoc analysis of STEVIE demonstrated that the number of treatment interruptions did not impact response rate with vismodegib treatment [124]. Multiple analyses also reported less severe AEs and longer treatment duration than continuous-dose studies when vismodegib dose interruptions were instituted [125,126,127]. Additionally, various vismodegib dose reduction regimens evaluated in patients with advanced BCC showed comparable treatment response rates and reduction in AE severity, though dose reductions for vismodegib are off-label [78, 120, 121, 126].

Several published case reports and case series have also demonstrated the success of treatment interruptions or dose reductions in mitigating AEs observed with HHI treatment without impacting efficacy of treatment [122, 128,129,130,131,132]. In the clinical setting, multiweek treatment interruptions or dose reductions in the form of alternate-day dosing may effectively help manage AEs with HHI treatment [12, 13, 20]. The NIELS study investigated real-world treatment patterns in patients with laBCC taking vismodegib and found that 53 of the 66 (80.3%) enrolled patients had documented treatment interruptions, with 35.8% of these reported to be due to unacceptable toxicity [133]. However, only one (1.5%) patient reported discontinuation due to AEs [133]. This suggests that, in clinical practice, HHI treatment is more often interrupted than permanently discontinued in response to AEs. Additionally, a retrospective observational analysis, which investigated the efficacy of interrupted dosing in managing AEs, found that patients experienced improvement of alopecia, muscle spasms, and dysgeusia upon switching from continuous to intermittent treatment [72]. These studies indicate that treatment interruptions can effectively manage AEs to prevent discontinuation in real-world practice. Following any necessary dosing modifications, supportive care interventions should be continued when restarting treatment [109].

6.4 Switching Hedgehog Pathway Inhibitors

Switching HHIs may be an option to help manage AEs for some patients, as described in multiple publications [134]. Reduced AE incidence was reported in several studies of patients with advanced BCC upon switching from vismodegib to sonidegib (± itraconazole) treatment [45, 72, 129, 135,136,137,138]. Another case report described improved tolerability in a patient with BCC switching from itraconazole to vismodegib treatment [139]. Interestingly, a real-world study investigating the safety and efficacy of HHI rechallenge in patients with recurrent BCC who previously achieved complete response (CR) with vismodegib found that the tolerability of vismodegib on rechallenge was better compared with that of the initial course. Thus, the improved tolerability of a second HHI course may be attributed to the treatment interruption itself, independent of the identity of the second agent [140]. Also of note, only 2 of the 35 patients in this study were switched from vismodegib to sonidegib for the rechallenge [140]. Given the similar mechanisms of action and safety profiles of HHIs, switching agents in patients experiencing class-effect AEs may not be an effective mitigation strategy for all patients [12, 61, 109]. More research is needed to confirm the potential benefits of switching HHI therapy in patients with BCC.

7 Conclusions

HHIs are considered a therapeutic breakthrough for patients with advanced BCC for which traditional therapies are not suitable [9, 109]. However, class-effect AEs, including muscle spasms, alopecia, and dysgeusia, can limit continuous treatment in patients with advanced BCC, which may, in turn, impact clinical outcomes [10,11,12]. Management of these and other AEs through interventions such as treatment interruptions and supportive care may improve tolerability and optimize treatment duration of HHIs for patients with advanced BCC who require long-term treatment [119].