Analysis of residual disease in periocular basal cell carcinoma following hedgehog pathway inhibition: Follow up to the VISORB trial

Abstract

Basal cell carcinoma (BCC) is a common skin cancer caused by deregulated hedgehog signaling. BCC is often curable surgically; however, for orbital and periocular BCCs (opBCC), surgical excision may put visual function at risk. Our recent clinical trial highlighted the utility of vismodegib for preserving visual organs in opBCC patients: 67% of patients displayed a complete response histologically. However, further analysis of excision samples uncovered keratin positive, hedgehog active (Gli1 positive), proliferative micro-tumors. Sequencing of pre-treatment tumors revealed resistance conferring mutations present at low frequency. In addition, one patient with a low-frequency SMO W535L mutation recurred two years post study despite no clinical evidence of residual disease. Sequencing of this recurrent tumor revealed an enrichment for the SMO W535L mutation, revealing that vismodegib treatment enriched for resistant cells undetectable by traditional histology. In the age of targeted therapies, linking molecular genetic analysis to prospective clinical trials may be necessary to provide mechanistic understanding of clinical outcomes. Trial Registration: ClinicalTrials.gov Identifier: NCT02436408.

Fig 1. Keratin positive micro-tumors present despite complete clinical response to vismodegib.

A) Pathology results for excision samples after vismodegib treatment (error bars indicate 95% CI, n = 9 (left column), n = 18 (right column)), B) Pre- and post-vismodegib histology, C) Average total area and D) Average lesion size (pixels) of keratin 5 positive staining (normalized to section length, error bars indicate SEM, unpaired t-test *p<0.05, n = 7 (left column), n = 11 (right column)), E) Clinical photo, histology, and keratin staining pre and post-vismodegib treatment of patients scored as “no sign of disease” (black arrow–tumor location, yellow box—area of peripheral palisading).

Fig 2. Persistent BCC expresses Gli1 and is proliferative.

A) Gli1 expression (dots/nuclei) pre- and post-vismodegib treatment in patient samples scored as disease present (red) or no sign of disease (green) (error bars indicate SEM, unpaired t-test *p<0.05, n = 7,6,9,11 columns left to right), B) Gli1 expression (brown dots) in excision samples from “disease present” (left panels) and “no sign of disease” (right panels) samples. C) Gli1 expression (left panels, brown) and proliferation (Ki67, green) in keratin 5 positive (red) persistent lesions from samples scored as disease present (top panels) and no sign of disease (bottom panels).

Fig 3. Mutational status of pre-vismodegib orbital BCC tumors.

A) Summary of all PTCH and SMO variants present in each tumor. Schematic of PTCH (B) and SMO (C) pathogenic variants present in patient cohort. D) Summary of variant data by patient, E) Classification of non-silent PTCH variants with >0.25 allele frequency, F) Classification of non-silent SMO variants.

Fig 4. Vismodegib treatment enriches for cells with resistant mutations.

A) Clinical photos and Gli1 expression (brown dots) in a patient pre- (left panels), post-vismodegib (center panels), and after recurrence (right panels), B) Allele frequency of PTCH1 and SMO variants pre-, post-vismodegib, and after recurrence (*variant not detected)(yellow star–predicted pathogenic variant, red star–known pathogenic variant).