Genetic alterations conferring resistance to hedgehog inhibitors in basal cell carcinoma

KEYWORDS:

• Hedgehog pathway
• basal cell carcinoma
• vismodegib
• sonidegib
• PD-1
• itraconazole

With great interest, we read ”Expert opinion on sonidegib efficacy, safety and tolerability” by Villani et al. The authors emphasize the promise of sonidegib and vismodegib, competitive inhibitors of the smoothened (SMO) membrane protein within the hedgehog (Hh) signaling pathway, which is implicated in the pathogenesis of various solid tumors including basal cell carcinoma (BCC) [1]. In the Hh pathway, the Hh ligand (Sonic Hh, Indian Hh, or Desert Hh) binds to the patched (PTCH1) protein membrane receptor, resulting in SMO inhibition. Downstream, if SMO fails to be inhibited, SMO inhibits suppressor of fused homolog (SUFU), resulting in translocation of glioma-associated oncogene (GL) transcription factors to the nucleus and activation of genes involved in tumorigenesis [1]. BCCs are typically slow-growing tumors confined to the skin that remain curable with surgery; however, rarely, they can invade underlying regional structures, a locally advanced BCC (laBCC), or exhibit distant spread to another organ or non-regional lymph node, a metastatic BCC (mBCC) [2]. laBCC and mBCC are forms of advanced BCC that present a therapeutic challenge as there may be complications associated with surgery or a high risk of recurrence following surgery [2].

Basal cell nevus syndrome, or Gorlin syndrome (GS), is due to mutations within the Hh pathway, and patients with GS typically present with multiple, and in some cases, more than 100 BCCs at the time of presentation [3]. GS most commonly involves mutations in PTCH1, but SUFU mutations may occur as well [3,4]. For these patients, it is difficult to surgically remove each BCC, and radiation therapy may be associated with a number of adverse effects, including cognitive deficit and seizure disorder [3]. Thus, SMO inhibitors are an appealing option for GS. While the use of SMO inhibitors may be effective for PTCH-1 associated GS, it is not effective for SUFU-associated GS because SUFU is downstream SMO in the Hh pathway [3–5]. Furthermore, infundibulocystic BCC is a subtype of BCC histologically distinguished by dermal proliferation of anastomosing cords and strands of basoloid cells with associated small infundibular-type, keratin-filled cysts [6]. Reports have noted SUFU mutations, which are downstream of SMO, in patients with infundibulocystic BCC [5,6]; therefore, this type of variant of BCC may be a clue to Hh inhibitor resistance. A case of infundibulocystic BCC with SUFU mutation showed no response to daily vismodegib treatment [5]. Thus, a potential challenge in the treatment of advanced BCCs with SMO inhibitors is resistance due to mutations downstream of SMO in the Hh pathway. Furthermore, inhibiting the Hh pathway in these resistant cases was also associated with considerable adverse effects [4,5]. Genetic testing of tumors may be indicated in cases of advanced BCC or infundibulocystic BCC to ensure that the existing Hh mutation occurs upstream of SMO to avoid unnecessary adverse effects of vismodegib or sonidegib [5].

Interactions between programmed death 1 (PD-1) and PD-1 ligand (PD-L1) are immune checkpoints that block the anti-cancer immune response, and monoclonal antibodies targeting PD-1 or PD-L1 led to improved survival in many cancers, including melanoma [7]. In cases of BCC in which Hh inhibitors are not appropriate, reports have demonstrated that pembrolizumab, a PD-1 inhibitor, may represent a therapeutic alternative [4,8]. In a study examining 40 BCCs, it was found that 82% (33 of 40) of samples demonstrated PD-L1 expression on tumor cells, and 22% (9 of 40) demonstrated PD-L1 expression on tumor infiltrating lymphocytes. PD-L1 was also observed in close geographic association with PD-1-positive tumor infiltrating lymphocytes [8]. Thus, PD-L1/PD-1 inhibitors may warrant further investigation in cases of Hh inhibitor resistance.

Another alternative to vismodegib or sonidegib is itraconazole, another Hh pathway inhibitor that has been demonstrated to inhibit SMO and GLI-1 [9]. Thus, it is theoretically able to block the Hh pathway even if SUFU is mutated. In mice, itraconazole was found to reduce GLI-1 mRNA levels in tumor cells in a mouse allograft model [10]. A phase II clinical study found that itraconazole reduced proliferation by 45% and GLI-1 activity by 65% in BCC [11]. However, this study was limited by its small sample size and the fact that most patients did not have advanced BCC, so further investigation is warranted [11].

Overall, BCCs are often diagnosed and treated and diagnosed promptly, with five-year recurrence rates found to be 2–3% [12]. Advanced BCCs comprise 1–10% of all cases, and mBCCs account for 0.028–0.5% of cases [12]. Thus, the rarity of advanced BCC has made it difficult to understand the optimal therapy [12]. In these cases, surgical excision may not be indicated due to proximity to vital structures, and SMO inhibitors such as vismodegib or sonidegib have shown promise and success in terms of efficacy, safety, and tolerability [1,12]. However, resistance to SMO inhibitors is a possibility, usually dictated by the location of the mutation within the Hh pathway. In cases of SMO inhibitor resistance, vismodegib and sonidegib are not effective, and their use may lead to unnecessary adverse effects. Alternatives such as immune checkpoint inhibitors or itraconazole warrant consideration in these cases.

Declaration of interest

None.

Additional information

Funding

This paper was not funded.