Background
Basal cell carcinoma (BCC) represents the most common (~80%) form of skin cancer worldwide in white people and its incidence is rising in many countries, although it is not systematically reported in tumor registries [
1]. Metastatic spread is extremely rare, but the incidence of locally advanced BCC (laBCC) has been estimated around 8/100,000/year and is associated with substantial morbidity, since most tumors occur in functional areas [
2,
3]. Excisional surgery and Mohs staged resection are the most effective treatments for low- and high-risk BCCs, respectively; radiotherapy, curettage and cautery, cryosurgery, carbon dioxide laser, photodynamic therapy, topical immunotherapy represent alternative options in selected cases [
4]. Long-term outcomes are crucial in evaluating BCC treatments since it is a slow-growing cancer and recurrence may take long time before being clinically apparent. During the last decade, electrochemotherapy (ECT) has become an appreciated locoregional therapy in the field of dermato-oncology. ECT exerts its anti-tumor effect through the permeabilization of cancer cells to chemotherapy by means of short, high-voltage, electric pulses which destabilize the cell membrane barrier, allowing intracellular access of chemotherapeutic drugs that otherwise would not be able to penetrate the cell effectively [
5]. Besides this permissive effect on chemotherapy, ECT exerts a complex vascular disrupting action, which may be usefully exploited when dealing with bleeding tumors [
6]. Drugs used in ECT, bleomycin or cisplatin, are cheap, easy to manage and generally safe. According to the European standard operative procedure of ECT (ESOPE), bleomycin can be administered either intratumorally or intravenously, according to the disease burden, while cisplatin can be injected intratumorally, in patients with few and small tumors [
7]. Since the publication of the ESOPE in 2006, an increasing number of studies have provided evidence of ECT efficacy on different tumor histotypes [
8‐
12]. On this basis, the National Institute for Health and Care Excellence (NICE) has recently recognized ECT as a safe alternative option for BCC patients (NICE interventional procedure guidance [IPG478], published date: February 2014). However, since the evidence basis is limited and established treatments provides high cure rates, clinicians involved in ECT are encouraged to systematically collect data on case selection, treatment parameters, patient outcomes and, possibly, to submit these data to the International Network for Sharing practices on Electrochemotherapy (InspECT) register (website:
http://www.insp-ect.org/) [
13]. Although BCC is by far the most frequent skin cancer, clinical experience with ECT is still scarce [
10‐
12,
14‐
18], probably due to the availability of several established treatment options [
19]. In the present study, we sought to examine the feasibility, efficacy and toxicity of ECT in BCC and to gain insights into this potential field of ECT application.
Discussion
As the prevalence of BCC rises with aging of the population, clinicians involved in BCC treatment face an increasing number of challenging scenarios. These are the patients with laBCC, hereditary basal cell naevus syndrome, and those with multiple or recurrent BCCs, who are compelled to undergo multiple interventions. Although these patients represent only a small proportion of the overall BCC population, nonetheless the burden of disease prompts for new evidence-based therapies especially for those individuals who are not amenable to surgical treatment (i.e., excision would be disfiguring or unpractical). In the present study, which is based on the largest available BCC series managed by ECT, we evaluated 84 subjects who were treated according to a standardized protocol (ESOPE) [
7]. Since ECT indications were restrictive, our study population has at least two peculiar characteristics: the high number of subjects with relevant comorbidities (Table
1), and the high percentage of laBCCs (49%). Since the presence of comorbidities has been shown to impact on the survival of elderly patients, especially for less lethal cancers [
22], a minimally invasive, low-demanding treatment modality can be considered a reasonable alternative in well-selected cases. The simplicity of treatment application, the ability to simultaneously treat multiple lesions (Additional file
4: Figure S4) coupled with high patient tolerability, represent potential advantages of ECT in patients with multifocal or laBCC or in presence of diseases that limit the number of practicable therapies [
17].
Electrochemotherapy was successfully applied in all cases, with no serious adverse events. Overall, we observed a 50% CR rate after a single ECT cycle, with significantly higher clearance rates in younger patients with primary BCC presentation, local tumor extent, small tumor size, well-defined borders, absence of ulceration and non-aggressive histology. Overall, local recurrence occurred in 20% of patients (primary BCC, 14%; recurrent BCC, 26%) and ECT-induced toxicity was observed in 19% of patients, being mild and transient in most cases. Noteworthy, we reported a significant decrease (from 42 to 18%) in the fraction of patients who required wound dressing during the follow-up. This is a relevant observation since BCC, and especially advanced BCC, is associated with a significant disease burden and health care resource utilization. The same parameters that were associated with CR achievement—with the exception of BCC presentation and TNM classification—were also associated with longer tumor control. Some of these variables support the notion of tumor response dependency upon tumor size in ECT, as previously observed by other authors [
8,
9,
23]. Interestingly, BCCs with ill-defined borders were less responsive to treatment. For this reason, we advocate the adoption and quantification of predefined treatment safety margins as a valuable measure to minimize the risk of recurrence from subclinical tumor deposits [
13]. In this regard, general anesthesia—together with no previous surgical treatment—was found to be associated with significantly higher CR rate in patients with non-melanoma head and neck skin cancers who underwent ECT according to a recent multicenter phase II study including 55 individuals, of whom 24 with BCC [
24]. This observation raises the hypothesis that general anesthesia may ensure better treatment tolerability by patients and thus a more accurate tumor coverage with electrode placements during the procedure.
In addition, in our study the patients with aggressive BCC histotypes proved to be less responsive; therefore, BCC-specific histopathological parameters should be taken into account in future studies in order to refine patient selection. Finally, we correlated treatment outcome with ECT procedural parameters, but these observations are likely biased by differences in BCC size [
7]. Accordingly, the identification of the best route of drug administration as well as the most suitable electrode type in ECT will require a dedicated study with more homogeneous tumor groups.
It should be noted that currently available ECT guidelines [
7] mainly refer to the feasibility of the procedure itself, but are not informative about the appropriateness of treatment indication or the optimal number of treatment cycles. These aspects need to be elucidated in future studies. According to our experience, a second ECT cycle increased the CR rate from 50 to 63% and retreatment was more advantageous in patients with local BCC, in whom CR rate increased from 72.5 to 85%. However, numbers are small and these findings should be evaluated with caution. For instance, the benefit of retreatment has been observed in BCC treated by 5-aminolaevulinic acid photodynamic therapy. In a longitudinal, non-randomized study on 44 patients with primary or recurrent BCC, Christensen et al. [
25] reported a CR rate of 60 and 87% after one or two treatment sessions, respectively. For the time being, retreatment with ECT seems a reasonable option in patients with small BCCs in order to consolidate response duration, particularly when risk factors such as aggressive histology, ill-defined borders or ulceration are present (Table
3).
Our results appear less satisfactory as compared to the literature data on BCC treatment with ECT [
24]. It is remarkable that no comparative trials have been carried out so far and no cost-analyses are available. The bulk of the literature consists of small observational series, often including heterogeneous skin cancers [
10,
12,
14‐
18,
24]. In an early trial, Glass et al. [
16] reported a 98% response rate after one or more ECT cycles with intralesional bleomycin and no recurrences detected after 18 months. However, this trial, which adopted a different electroporation protocol, enrolled exclusively patients with sporadic BCC and low tumor size (mean diameter 0.91 cm, range 0.37–2.1). A recent meta-analysis assessing 47 prospective studies on patients with cutaneous metastases—mainly from melanoma and breast cancer—treated by skin-directed therapies (including ECT, radiation, photodynamic, intralesional and topical therapies) showed sustained response rates across treatments with G ≥ 3 dermatological toxicity in less than 6% of patients [
26]. More recently, a European and an Italian multicenter studies (both adopting the ESOPE protocol), have reported CR rates of 91 and 66.7%, respectively, in BCC patients, although with a short follow-up (6 and 13.9 months, respectively) [
10,
11]. In 2016, Rotunno et al. reported a 75% CR rate in patients with head and neck BCC treated by ECT. The majority (42%) of these tumors (whose median size was 24 mm) were located on the scalp and half of patients required at least 2 ECT cycles [
24]. An ongoing randomized trial is currently evaluating treatment durability of surgical excision versus ECT in patients with primary BCC. An interim analysis, based on 86 patients at 3-year follow-up, indicates comparable efficacy, as indicated by local disease-free progression of 97 and 92%, respectively (
p = 0.37) [
27].
Recurrence rates in primary and recurrent BCC treated by conventional treatments can be summarized as follows: excisional surgery, 2–17.4% [
28‐
31]; Mohs surgery, 1.0–5.6% [
28,
32]; curettage and electrodessication, 7.7–40% [
28,
33]; cryotherapy, 7.5–13% [
28,
32]; radiotherapy, 7.4–16% [
28,
32,
34,
35]. Topical immunotherapy with imiquimod ensures 5-year recurrence rate around 20% in small superficial BCC [
36]. Finally, recurrence rate after photodynamic therapy, in its various forms of ALA-PDT (aminolevulinic acid photodynamic therapy) and MAL-PDT (methyl aminolevulinate photodynamic therapy), ranges to 18–22% [
37,
38]. Taken together, these data indicate that ECT has an acceptable toxicity profile, but a higher recurrence rate, especially when compared with excisional and Mohs surgery. In the present study, 5-year recurrence rate in local and laBCC was 20 and 38%, respectively. There are same possible explanations for these findings. First, we cannot exclude suboptimal treatment delivery. ECT exerts its effect when tumors are simultaneously exposed to chemotherapy and electric fields. In this regard, BCC poses peculiar challenges. More than 60% of our patients had tumors located on the head and neck region, where direct drug injection can lead to inhomogeneous distribution or to spillage from the injection site, as previously observed [
39]. Further, we hypothesize that the diffusion of chemotherapy in tumor tissue may have been hampered by the presence of scars from previous treatments, thus leading to inhomogeneous drug distribution [
24]. Second, it is possible that some tumors could have been covered sub optimally by electric pulses. Since 36.9% of patients were managed under local anesthesia only, it is likely that sedation or general anesthesia would have allowed a deeper (e.g. in laBCCs) or wider (e.g. in BCCs with ill-defined borders) electrode application [
24]. In support of this hypothesis, it is worthy to note that in the present series 20 out of 84 patients had tumors with clinically ill-defined borders and that these tumors had the highest recurrence rate (Table
3). A third explanation may reside in tumor biology. In fact, the fraction of replicating cells in BCC is low [
40] and cytotoxic agents, which selectively kill the actively replicating cells, exert their action on a small fraction of them. Finally, the relatively long follow-up of our study allowed for the detection of late recurrences.
Surgical treatment definitely represents the preferred options for BCC, with conventional excision being sufficient in most primary BCCs and Mohs micrographic surgery being the most effective in high risk or recurrent BCCs of the face, where it is associated with 5- and 10-year recurrence rates of 2.1–5.2% and 3.9–4.4%, respectively [
41‐
43]. Among non-surgical options, radiotherapy represents a consolidated and effective alternative [
3,
4,
34,
35], although logistical barrier may be a limitation. In selected cases, ECT may be a rapid, easy to apply treatment, although the broad spectrum of BCC presentation poses peculiar challenges to its application. In order to maximize the efficacy to toxicity ratio, clinicians should aim to select the most appropriate ECT treatment modality (type of anaesthesia, route of drug administration and electrode geometry). In theory, systemic chemotherapy, coupled with use of a large needle electrode (i.e., the hexagonal array, Fig.
1a) allows homogeneous tumor tissue exposure to chemotherapy and complete tumor electroporation. However, this should be weighed against possible side effects. Systemic bleomycin can be associated, although in very rare cases, with lung toxicity (especially in patients >70 years of age) [
44], and the application of a large needle electrode can increase skin injury due to its greater invasiveness [
39]. Furthermore, clinician and patients could be reluctant to systemic chemotherapy for the treatment of a localized, slow growing tumor. ECT proved to be effective in BCCs located on the midface, where treatment could be potentially challenging due to the presence of aesthetic and functional structures. According to a well-documented series of three patients with BCC of 0.5–1.0 cm
2 affecting the peri-ocular region, treatment with ECT is feasible, safe and associated with acceptable scarring [
18].
Interestingly, we observed a 31.7% CR rate in laBCC (Table
3), with appreciable tumor shrinkage also in partial responders (Additional file
5: Figure S5). Only a few studies of chemotherapy in BCC have been published and most of them found a relative unresponsiveness, except for limited success with cisplatin [
45]. We did not perform cisplatin-based ECT, since, according to the ESOPE, this drug is currently codified for intratumoural injection only, and specifically in small size tumors [
7]. Although the laBCC field has successfully entered the era of targeted therapy, locoregional treatments continue to be worth of consideration, particularly in the frame of integrated strategies. CR rates ranging from 21 to 34% have been consistently observed with the hedgehog inhibitor vismodegib, with median PFS ranging from 9.5 to 24.5 months [
46,
47]. However, tolerability still represents a tangible obstacle in continuation of therapy, with 10–25% of patients requesting to stop treatment and 36% discontinuing due to adverse events (mainly muscle spasms, alopecia and dysgeusia) [
46,
47]. In this context, local treatment with ECT could be rationally associated or combined (e.g. during the so-called
drug holidays) with target therapy in order to maximize or consolidate tumor response. However, for the time being, the limited experience does not allow general recommendations, therefore prospective trials are warranted to assess the impact of ECT in advanced BCC.
Although non-comparative, the present study has some strengths. It is based on a relatively large series. It included BCC-specific parameters (histological subtypes, tumor borders, ulceration) in order to personalize treatment indication. It was conducted at a center, which is long acquainted with the procedure [
8,
9,
11,
39]. Finally, it has a relatively long follow-up, which is crucial for assessing late recurrences [
28,
43]. Nevertheless, we also acknowledge a number of limitations. First, this study is not comparative. Second, we included patients with heterogeneous BCC types and adopted a conservative definition for laBCC. Third, tumor response was clinically assessed, and thus possibly overestimated, with no independent review assessment. Fourth, we did not apply treatment safety margins around the target tumor in all cases. Finally, patient-reported outcomes and functional as well as cosmetic results were not included. It is desirable that disease-specific questionnaires will be incorporated in future studies together with 5- and 10-year follow-up data.