Introduction

Vismodegib (Erivedge, Genentech) is a first-in-class inhibitor of the hedgehog (Hh) pathway, which received approval from the federal drugs authority in the United States of America in January 2012. It was licensed for both locally advanced basal cell carcinoma (BCC) not amenable to surgical intervention, and for metastatic BCC [1,2,3]. Vismodegib acts via direct inhibition of the protein-coupled receptor smoothened, and inhibits downstream signalling pathways involving transcription factors such as Gli1 [1].

In 2017, the National Institute for Health and Care Excellence (NICE) withdrew its recommendation for the use of vismodegib for the treatment of BCC, citing a lack of survival data in locally advanced disease, and inadequate data comparing outcomes to standard supportive care, with vismodegib having a significantly higher cost per quality-adjusted life year. However, vismodegib continues to be widely used in other countries for cases of locally advanced and metastatic BCC. It has also been shown to have utility in locally advanced periocular BCC [4,5,6,7,8,9,10].

The purpose of this multicentre, international case series is to report outcomes of patients with locally advanced periocular BCC who have been treated with vismodegib.

Patients and method

A retrospective review of clinical case notes was performed for all patients by the identifying institution. All cases of vismodegib treatment for periocular BCC in the following institutions were included in this study: The Sussex Eye Hospital (Brighton, UK), Manchester Royal Eye Hospital (Manchester, UK), Cardiff Eye Unit (Cardiff, UK), Newcastle Eye Centre (Newcastle, UK), Queen Victoria Hospital (East Grinstead, UK), Wellington Hospital (New Zealand), and the Royal Brisbane and Women’s Hospital (Brisbane, Australia). Baseline data for each patient included age, sex, race/ethnicity, BCC subtype, primary location and laterality, primary or recurrence, presence of orbital extension, presence of metastatic disease and ophthalmic signs (vision/motility/lacrimal apparatus involvement).

The dose, duration, compliance, response to treatment and side effects of vismodegib treatment were collected, as well as prior and adjuvant and post-vismodegib surgical treatment and histopathological findings. A complete response to treatment was defined as complete regression of the tumour. A partial response was defined as regression of tumour but not to the extent of a complete response.

The Human Research Authority online decision-making tool determined that ethics approval was not required for this retrospective review.

Results

Demographics

Thirteen patents were identified in this retrospective review. Nine patients were from UK institutions, three from the Royal Brisbane and Women’s Hospital (Australia), and a further one patient from Wellington, New Zealand.

Seven patients were male and six female. The mean age was 75 (median 76, range 43–91). All patients were Caucasian (Table 1).

Table 1 Demographic data, treatment, response to treatment, side effects and follow-up of patients treated with vismodegib for periocular BCC.

Basal cell carcinoma characteristics

Seven patients had disease centred on or around the medial canthus. Three patients had disease centred on or around the lower eyelid, and one patient had BCC on both the upper and lower eyelid. One patient had disease arising from the forehead and extending down to the superior orbit, and a further patient had disease centred around the right temple region extending to the lateral canthus.

All lesions were ill-defined with 5/13 being described as ulcerating. All lesions were biopsy-confirmed BCC with histopathological subtypes of infiltrative (4), nodular (3), nodular/infiltrative (1), micronodular/infiltrative (1), baso-squamous (1) (BSCC), superficial (1) and cystic (1) and one not available (NA).

In 8/13 cases (62%) the BCC was recurrent after previous surgery and the disease was primary in the other five cases. One patient had basal cell nevus syndrome.

Orbital involvement was demonstrated in seven patients (58%) on magnetic resonance imaging. No patients had metastatic disease.

Co-existing ophthalmic signs

The best corrected visual acuity was mildly reduced in 4/13 (31%) patients and one had no perception of light vision on the affected side. The ocular motility was impaired in 5/13 (42%), all of whom had orbital involvement. The lacrimal apparatus was involved in 4/13 cases (38%).

Vismodegib treatment

All patients received a vismodegib dose of 150 g once daily. Compliance was reported as >90% in eight cases, with data for the remaining five unavailable. One patient had to stop treatment due to side effects after 2 months.

The mean treatment time with vismodegib was 7 months (range 2–36 months, median 7 months).

Side effects

Eleven out of 13 patients (85%) reported side effects secondary to treatment. Reported symptoms were: fatigue (6/13), dysgeusia (4/13), weight loss (3/13), reduced appetite (2/13), muscle cramps (2/13), alopecia (2/13), flatulence (1/13).

Prior treatment

Eight out of 13 patients (62%) had received prior treatment for BCC before commencing vismodegib therapy. Seven of these eight had undergone excision, of which three also had adjunctive radiotherapy. One patient had received cryotherapy alone.

Outcomes

Follow-up

The patients were followed for between 12 and 48 months (mean 30 months, median 24 months).

Response to vismodegib

A complete response to treatment was reported in 5/13 (38%) patients and a partial response in eight (62%) patients. In one of the eight patients with a partial response initially, progression was then seen and the patient underwent radiotherapy.

Subsequent surgical treatment

Out of 13 patients, six patients had tumour excision surgery following the initial course of vismodegib (such as Fig. 1). Three patients underwent exenteration surgery; in one of these the procedure was conducted on cessation of vismodegib, because the tumour had not regressed sufficiently to alter the original surgical plan. In the other two globe-sparing surgery was initially undertaken, but exenteration was subsequently conducted for later recurrence. The remaining three patients had cutaneous excisions, of whom one had had orbital involvement prior to vismodegib treatment but significant regression facilitated pre-septal surgery.

Fig. 1: Example of tumour regression following course of vismodegib and subsequent resection and reconstruction with full thickness skin graft.
figure 1

Male patient with medial canthal/bridge of nose BCC at presentation (a), treated with vismodegib showing partial response (b). Subsequent resection and reconstruction with no recurrence at 2 years (c).

Recurrence

Recurrent BCC occurred in 3/13 (23%) patients after either vismodegib treatment alone or vismodegib treatment with post-treatment surgical excision. One of these patients developed recurrence on the sclera following globe-sparing surgery and loco-regional flap reconstruction, and subsequently underwent exenteration. It is unclear in this patient whether the recurrence was from the original visible tumour site (medial canthus) or from a deep seated focus of tumour. One patient demonstrated an apparent complete regression of the tumour from vismodegib treatment, but then developed a small recurrence on the forehead (original primary site) necessitating excision and reconstruction with a full thickness skin graft. The third patient had 3 months of vismodegib treatment followed by local resections using Mohs micrographic surgery with apparently clear margins, but subsequently developed recurrence. This patient had a further 2 months of vismodegib treatment, but there was minimal response, so underwent exenteration.

Discussion

In this study we report a partial or complete response to vismodegib in all 13 patients with locally advanced periocular and orbital BCC, and in five cases the treatment obviated the need for surgical excision of the tumour. However, 11 patients suffered significant side effects from the vismodegib treatment and exenteration was still required in three patients.

There is an increasing body of evidence on the outcomes of vismodegib for non-periocular BCC with studies reporting between 60% partial response rates up to 68.5% complete response rates [1,2,3]. There is a paucity of data on the outcomes of vismodegib treatment for locally advanced periocular BCC. Small prospective and retrospective series report complete response rates between 25 and 67% with only 13% showing no response in one study [5,6,7]. In the present series 38% (five patients) demonstrated a complete response to vismodegib, with the rest demonstrating a partial response.

Prolonged vismodegib treatment may cause significant side effects and necessitate cessation of treatment. The STEVIE study (The SafeTy Events in VIsmodEgib study) reported muscle cramps (in 7% of patients), dysgeusia (5%), weight loss (4%), alopecia (3%), asthenia(3%), fatigue (2%), ageusia (2%) and nausea (1%) [3]. Data from the ERIVANCE trial suggested a higher incidence of adverse events, particularly muscle spasm, when treatment lasted more than 12 months. Hepatotoxicity which has been described in association with vismodegib was not seen in any of the patients in this series. It is not clear why the side effect profile was different in the present series in which fatigue was reported by nearly half the patients and dysgeusia in nearly a third (but only one patient had to stop treatment) but it may reflect the long duration of treatment (mean 10 months) or increasing awareness of potential side effects and subsequent specific questioning of patients.

Surgery is the mainstay of treatment for BCC with orbital invasion. Exenteration is frequently required and is a disfiguring operation that carries significant surgical and psychological morbidity [11, 12]. Adjunctive radiotherapy may be beneficial in some cases, though may also cause sight threatening ocular surface disease negating the major benefit of globe-sparing surgery and may hinder socket recovery after exenteration. Globe-sparing surgery can be utilised, but is still a major surgical intervention that can result in intractable diplopia, epiphora, and lid malposition and may not achieve complete excision margins [13]. In the present study, vismodegib treatment achieved sufficient tumour reduction in 3 of 13 patients enabling globe-sparing excision surgery instead of the anticipated exenteration. In a further patient, the apparent reduction in tumour size from the vismodegib was falsely reassuring as globe-sparing surgery was attempted, but clear margins were not attained, and subsequent recurrence on the sclera necessitated exenteration.

Despite some of the beneficial responses in this and other series, there remain areas of great uncertainty in vismodegib treatment combined with huge variation in individual treatment response. For example evidence is lacking for (1) the optimal treatment duration, (2) how to assess treatment response, (3) predictors for the degree of response to treatment of individual patients (4) which tumours will truly reduce in size throughout their margin and which will showed patchy reduction leaving tumour foci throughout the original area that may be harder to delineate and excise (5) to what degree tumours develop treatment resistance [14, 15].

This is a retrospective study and may be affected by recall bias, although the small number of patients treated in each unit combined the complexity and extended nature of treatment may reduce this. In addition, patients in this series come from two continents. Whilst all patients included in this study were Caucasian, it is possible that environmental factors, such as ultraviolet exposure or genetics may influence the nature of the tumour and its response to treatment.

This retrospective review demonstrates vismodegib to be an effective, safe and well-tolerated adjunctive treatment for advanced periocular and orbital BCC. It may obviate the need for radical and disfiguring surgery such as orbital exenteration in selected patients.

Summary

What was known before

  • Vismodegib has previously been shown to have utility in periocular BCC although data regarding outcomes is limited.

  • Recommendation for use of vismodegib was recently withdrawn by NICE due to a lack of evidence regarding versus standard supportive care.

What this study adds

  • Vismodegib appears to be of use in selected patients and may obviate the need for major and disfiguring surgery such as exenteration.

  • Vismodegib appears to be well tolerated in this group of patients with fatigue being the most common side effect.