Introduction

Melanoma is the most aggressive form of skin cancer, representing 4% of all dermatological cancers. A total of 96,480 new cases of melanoma will be diagnosed, and 7230 deaths are expected in 2019 in the USA (Siegel et al. 2019). Its etiology is multifactorial, being associated with both environmental and genetic factors. BRAFV600E is the most common mutation, being observed in around 50% of melanomas and contributing to tumor growth, angiogenesis, and metastatic progression (Garnett and Marais 2004; Haass et al. 2004). Current treatments for advanced melanoma include chemotherapeutic and immunotherapeutic strategies, such as dacarbazine (DTIC), vemurafenib, interferon, interleukin 2 (IL-2), imiquimod, and checkpoint inhibitors. In addition, a number of recent studies are identifying new therapeutic strategies and targets for reducing melanoma’s metastatic potential (Orgaz and Sanz-Moreno 2013; Mattia et al. 2018). One of these treatments is Mycobacterium bovis Bacillus Calmette–Guerin (BCG). BCG is an attenuated strain that has been used as an immunotherapeutic agent for melanoma and superficial urothelial carcinoma (Begnini et al. 2015; Maruf et al. 2016). Nowadays, wild-type BCG and recombinant strains are used in combination with chemotherapeutic and immunotherapeutic agents to enhance the immune response and tumor regression (Stewart and Levine 2011) (Fig. 1).

Fig. 1
figure 1

Administration of wild-type or recombinant BCG strains alone or in combination with a chemotherapeutic or immunotherapeutic agent enhances immune responses and tumor regression in melanoma patients

Full size image

An improved understanding of the biological, genetic, molecular, and immunologic factors contributing to the progression of metastatic melanoma (Brandner and Haass 2013; Griewank et al. 2014; Shtivelman et al. 2014) may help identify novel therapeutic strategies (Ribas et al. 2011; Sullivan and Flaherty 2014). In addition, advanced proteomic technologies may be useful for detection and analysis of proteins involved in melanoma progression (Findeisen et al. 2009; Bougnoux and Solassol 2013). In this review, we highlight several studies utilizing BCG immunotherapy in combination with other therapeutic agents for the treatment of metastatic melanoma.

Melanoma

Melanoma is a neoplastic disorder caused by the malignant transformation of normal melanocytes and represents the most aggressive type of skin cancer (Bandarchi et al. 2010). Most cases originate in the skin, followed by the eyes and mucous membranes. Melanoma is associated with environmental factors and patient demographics, such as lighter skin tone, sun sensitivity, presence of atypical nevi, multiple freckles, people with weakened immune systems, and family history of melanoma (Gallagher et al. 2005; Bishop et al. 2007). Its increasing incidence is due to different ethnicity, geographical location, and excessive exposure to ultraviolet radiation combined with its high metastatic potential which has resulted in a significant increase in mortality (Gallagher and Lee 2006; Ali et al. 2013; Arnold et al. 2018).

Melanoma is subdivided into four different types: superficial spreading melanoma (70% of melanomas), nodular melanoma (15–30%), acral lentiginous (5–10%), and lentigo malignant (5%) (Goldstein 2001; Bastian 2014).

Deregulation of the MAPK, PI(3)K-AKT, P16INK4A/Rb), and Wnt/b-catenin signaling pathways is observed in metastatic melanoma and has been implicated in its etiopathogenesis and invasive behavior (Orgaz and Sanz-Moreno 2013; Alegre et al. 2014; Gurzu et al. 2018). Somatic mutations in the BRAF gene are observed in 40–60% of melanoma cases (Curtin et al. 2005; Abildgaard and Guldberg 2015). The most commonly observed mutation results in the substitution of lysine for valine (V600K mutation) (Long et al. 2011; Lovly et al. 2012) and is responsible for the activation of the MAPK pathway, which regulates normal cell growth and survival (Shinozaki et al. 2007; Flaherty and McArthur 2010; Roskoski 2010). Other genes commonly mutated in melanomas include NRAS (15–30%) (Gorski et al. 2005), CDK4, CDKN2A (observed in 20 to 40% of families with melanoma susceptibility) (Puig et al. 2005; Potrony et al. 2015), p16, PTEN, AKT1, MAP2K1, MAP2K2, MAP3K5, and MAP3K9. In addition, a number of DNA methylation changes have been reported (Hodis et al. 2012; Nikolaev et al. 2012; Stark et al. 2012). These mutations and epigenetic alterations trigger and promote the secretion of growth factors that contribute to cell proliferation, angiogenesis, alterations in the extracellular matrix, cytoskeletal organization, and metastasis (Haass et al. 2004; Orgaz and Sanz-Moreno 2013) to various organs such as the lungs, liver, brain, and bones (Leong 2003).

Melanoma diagnoses are made by dermatoscopy examination and advanced digital computer imaging techniques following the morphologic features summarized by the asymmetry, border, color, diameter, and elevation (ABCDE), which are confirmed by biopsy and histopathological examination. This process also allows for melanoma staging and prognostic, as well as TNM (tumor, node metastasis) classification for detection of distant metastases (Abbasi et al. 2004; Balch et al. 2009; Ciudad-Blanco et al. 2014; Lattanzi et al. 2019).

Therapies for metastatic melanoma treatment

Treatment for metastatic melanoma (MM) is mostly based on systemic therapy, although radiotherapy and surgical treatments are also used. However, current treatments have not resulted in significant improvements in patient survival due to adverse effects (Smith et al. 2007). Surgery is mostly used for resection of distant metastases (Luther et al. 2019). Radiation therapy can be used as adjuvant therapy for systemic therapy; however, some patients with symptomatic metastases can benefit from radiotherapy (Bhatia et al. 2009).

The first systemic therapy used for MM treatment is the cytotoxic chemotherapeutic agent Dacarbazine (DTIC). DTIC is an alkylate agent and is the first chemotherapeutic agent approved for MM by the US Food and Drug Administration (FDA) in 1975. Another treatment that had been approved for MM treatment in the USA is interleukin-2, an immunostimulatory cytokine involved in T cell proliferation. Interleukin-2 trials have demonstrated a 15 to 20% response rate for MM patients (Schwartzentruber et al. 2011); however, this drug failed to demonstrate survive prolongation (Garbe et al. 2011). Other cytokines being used to treat MM include interferon-alpha (Luther et al. 2019), the first cytokine to demonstrate activity in MM, with tumor response ranging from 10 to 20% (Schadendorf et al. 2009).

Since 2011, ten new agents have been approved by the FDA, including targeted therapies, immunotherapies, cancer vaccines, and other small molecules that can act as monotherapies or in combination (Simeone and Ascierto 2017; Luther et al. 2019). Ipilimumab is an IgG1 monoclonal antibody targeting cytotoxic T lymphocyte antigen 4 (CTLA-4). Treatment with ipilimumab was the first treatment to demonstrate a survival advantage for patients with MM (Wolchok et al. 2010). Other checkpoint inhibitors approved for MM include the programmed death 1 (PD-1) inhibitors nivolumab and pembrolizumab (Albertini 2018). These PD-1 inhibitors were first approved for patients with melanoma refractory to vemurafenib and/or ipilimumab. Atezolizumab, avelumab, and durvalumab are PD-L1 inhibitors that have also been approved for use in MM patients with advanced melanoma (Arulananda et al. 2018; Callahan et al. 2018; Hogan et al. 2018).

Talimogene laherparepvec (TVEC) is the only oncolytic virus approved for melanoma treatment (Luther et al. 2019). Another immunotherapeutic agent is allovectin-7 (velimogene aliplasmid) which is well tolerated, reduction in tumor size, and seems to be safe for the treatment of stage III or IV melanoma (Gonzalez et al. 2006; Bedikian et al. 2010; Sloot et al. 2016).

Drugs targeting altered mitogen-activated protein kinase (MAPK) pathway signaling—commonly disrupted in MM due to BRAFV600E mutations—have been approved form MM treatment, including dabrafenib and vemurafenib (Heakal et al. 2011; Hauschild et al. 2012). Other approaches to treat MM include combination therapies. Combination therapies can be a solution for the treatment of MM that is resistant to individual therapies (Srivastava and McDermott 2014; Gazzé 2018). However, one potential problem regarding the use of combination therapies is increased toxicity. The combination of vemurafenib and cobimetinib, dabrafenib and trametinib, or encorafenib and binimetinib have become standard treatments for MM patients carrying the BRAFV600E mutation (Ribas et al. 2016). Immunotherapies can also be combined. The most well-established and studied combination is ipilimumab and nivolumab (Hodi et al. 2016).

Despite recent advances, these therapies still have limitations including low response rates, adverse side effects, and resistance, especially for anti-PD-1/PD-L1 checkpoint inhibitors (Broussard et al. 2018). For these reasons, novel treatment strategies for MM are currently being developed (Olszanski 2014; Arulananda et al. 2018; Sullivan et al. 2018) to improve targeted therapy through identification of new targets for MM and combine current and future drugs (Mitchell 2003; Finn et al. 2012; Broussard et al. 2018).

Bacillus Calmette–Guérin

BCG is an attenuated Mycobacterium bovis strain developed by Albert Calmette and Camille Guérin that is widely used as a vaccine to protect against tuberculosis (TB) (Hart and Sutherland 1977; Dietrich et al. 2003) and has also been used as the first-line immunotherapy for intravesical treatment of superficial urothelial carcinoma (Ahn et al. 2014; Begnini et al. 2015; Maruf et al. 2016), resulting in reduced recurrence and progression (Kresowik and Griffith 2009; Askeland et al. 2012; Morales et al. 2015; Donin et al. 2017). However, BCG administration can cause adverse side effects including hematuria, irritation, local inflammation, and cystitis (Koya et al. 2006). The exact mechanism of action is unknown but it involves BCG binding to the fibronectin of the bladder wall followed by its internalization via macropinocytosis and presentation of BCG antigens to T cells (Lattime 1992; Zhao et al. 2000; Redelman-Sidi et al. 2013), causing tumor cell killing through signaling via the Toll-like receptors (TLR) 2, 3, 4, and 9, resulting in cytokine secretion and induction of the local inflammatory response (Miyazaki et al. 2006; Suttmann et al. 2006; Naoe et al. 2007).

There are several BCG modifications being used for immunotherapy and vaccination (Yuan et al. 2010; Zheng et al. 2015). BCG is sub-classified in different strains (Leung et al. 2008; Hayashi et al. 2009; Liu et al. 2009), and BCG exhibits anti-proliferative activities and results in the production of cytokines including IL-6 and IL-8 (Secanella-Fandos et al. 2013). Also, BCG has been genetically manipulated to secrete recombinant proteins (Borsuk et al. 2007; Begnini et al. 2013; Oliveira et al. 2017), foreign antigens from parasites, bacteria, and viruses (Bastos et al. 2009), and pro-inflammatory cytokines including IL-2, IL-8, and IL-18 that can enhance humoral and cellular immune responses (Biet et al. 2002; Luo et al. 2003; Luo et al. 2004).

rBCG::Ag85B-IFN-γ has been used in C57BL/6 mice to enhance immune responses and induced tumor necrosis factor TNF-α and IFN-γ expression (Liu et al. 2017). In addition, recombinant BCG::Rv2645 has been shown to improve dendritic cell (DC) antigen presentation and enhance Th1/Th17 immune responses against tuberculosis (Luo et al. 2018). The pantothenate auxotroph strain of Mycobacterium bovis BCG (BCGDpanCD) expressing HIV-1 Gag, Gp120, and RT induces a production of IL-2 and TNF-α by T cells (Chapman et al. 2013). Another recombinant BCG strain (ΔureC::hly) expressing a heterologous protein of Listeria monocytogenes has been shown to trigger production of caspases, IL-18, and IL-1β (Saiga et al. 2015). A similar study with ΔureC hly+ BCG was reported by Desel et al. (2011) and some vaccines with this construction are being tested for clinical efficacy to demonstrate their immunogenicity and safety (Nieuwenhuizen et al. 2017).

As an alternative to BCG, Mycobacterium vaccae (M. vaccae) has been shown to be safe, well tolerated, and results in long-term survival in patients with MM (Cananzi et al. 2013). In addition, the use of Mycobacterium indicus pranii (Mw) for the treatment of melanoma is being studied due to its ability to inhibit the matrix metalloproteinase (MMP-9), inhibiting melanoma tumor cell growth, and reducing invasiveness and metastatic potential (Halder et al. 2017).

BCG combination in melanoma treatment

BCG is used as an immunotherapeutic agent for the treatment of cutaneous MM. It is commonly administered alone or in combination with an autologous tumor cell vaccine or drug by intralesional injection (Sloot et al. 2016) for different stages of melanoma, resulting in regression of local and regional tumors and improved patient survival (Lotem et al. 2002; Triozzi et al. 2011). There are many different BCG-based combination therapies currently in use for the treatment of melanoma (Table 1). One of the most common combinations is BCG with DTIC which results in a 36.5% increase in the 10-year survival rate for stage II melanoma patients (Cascinelli et al. 1989). Additional combinations such as BHD (hydroxyurea with DTIC), BHD plus BCG, and DTIC plus BCG have been tested in patients with disseminated MM, resulting in overall response rates of 31%, 27%, and 18%, respectively (Costanzi et al. 1982). Even, the use of BCG and retinoids such as vitamin A palmitate and tretinoin have been shown to display antitumor activity for MM (Meyskens Jr. 1982).

Table 1 Combination BCG treatments for melanoma
Full size table

In a study for patients with stage IV melanoma, the use of the allogeneic cancer vaccine canvaxin™ in combination with BCG following complete resection resulted in antitumor immune responses and a 5-year overall survival rate of 39% compared with that of 20% for unvaccinated patients (Hsueh et al. 2002). This combination also increased the number of cytokine-producing CD4+ T cells, which correlated with overall survival (Hsueh et al. 2004). Faries et al. (2017) performed the same study on 246 stage IV melanoma patients treated with canvaxin™ and BCG following complete surgical resection and compared the results with 250 patients treated with BCG/placebo administered by intradermal injection. These results showed no improvement in patient outcomes and a 5-year survival rate of 34.9 months for patients treated with canvaxin™ and BCG compared with that of 39.1 months for patients treated with BCG/placebo. A similar study using BCG in combination with DTIC reported that this combination caused mild toxicity and was well tolerated, but did not improve patient outcomes compared with BCG alone (Agarwala et al. 2004).

Polyvalent vaccines have been shown to increase IgM responses and enhance the humoral immune response in patients with stage II melanoma, resulting in improved survival (DiFronzo et al. 2002). CancerVax has also been shown to induce an increase in TA90-IC antibodies in patients with advanced melanoma resulting in prolonged survival. TA90-IC is an antigen expressed on melanoma cells that serve as a prognostic marker, with higher TA90-IC levels present in patients with stage IV melanoma (Tsioulias et al. 2001). Another study reported immunization of MM patients with either smallpox (vaccinia) or BCG, showing that the vaccination reduced the risk of death and prolonged patient survival (Kölmel et al. 2005).

Additionally, the administration of BEC2, an anti-idiotypic mouse monoclonal antibody that mimics GD3 ganglioside mixed with BCG (BEC2-KLH/BCG) for patients with stage III and IV melanoma resulted in a 78% survival rate over 28 months (Yao et al. 1999). Furthermore, the GD3 ganglioside can be used as a target for melanoma immunotherapy, showing that 10 patients (42%) developed a detectable anti-GD3 antibody response, but did not correlate with survival outcomes (Chapman et al. 2004). It has been reported that BCG is a therapeutic option for stage III in-transit melanoma and can induce chemokine expression and regression of skin lesions (Yang et al. 2017). Another study developed a therapeutic vaccine named CSF-470 for cutaneous melanoma using a mixture of cell lines, BCG, and granulocyte-macrophage colony-stimulating factor (GM-CSF) as adjuvants in phase II–III trials, which resulted in the recruitment of Ag-presenting cells and antitumor immune responses (Aris et al. 2015). Other studies have demonstrated the use of a M-VAX vaccine composed of autologous melanoma cells and BCG that induces interferon-gamma-mediated T cell triggering inflammatory responses (Carretero et al. 2008).

The use of autologous melanoma vaccines and BCG plus recombinant human GM-CSF (rhGM-CSF) for treatment of stage IV melanoma has been shown to result in a complete response in 10% of patients with a median survival of 232 days (Leong et al. 1999). One such vaccine, VACCINEL, in combination with BCG and rhGM-CSF was administrated to patients with stage IIB/IV melanoma, resulting in well-tolerated, mildly toxic effects and increased antitumor immune responses (Barrio et al. 2006). Similar studies using the CSF-470 vaccine in combination with BCG and rhGM-CSF were well tolerated (Mordoh et al. 2017) and induced the release of pro-inflammatory cytokines such as IL-1β and TNF-α, stimulating long-term cellular and humoral immune responses (Pampena et al. 2018). On the other hand, Lotem et al. (2016) observed that the use of an autologous melanoma vaccine conjugated to dinitrophenyl and mixed with BCG in patients with stage III melanoma, followed by administration of ipilimumab, increased the response rate to 46% compared with 19% of patients treated with ipilimumab alone.

The continued generation of recombinant fusion proteins for use in melanoma immunotherapy is important as MUC1 and maltose-binding protein (MBP) in combination with BCG induces a specific cellular immunity, synergistic activities, NK cell activity, and increased expression of IFN-γ and IgG2c (Fang et al. 2014).

BCG in melanoma immunotherapy

Recently, immunotherapeutic strategies have been used in cancer therapy to activate the immune system and destroy cancer cells. While the antitumor mechanism of BCG has not been clearly elucidated, BCG has been used as a nonspecific immune stimulant to induce long-lasting immune responses (Wada et al. 1996). Indeed, different tumor immunization strategies and molecularly targeted therapies against the MAPK pathway are being studied to improve the response rate and survival for advanced melanoma patients (Davar et al. 2012). However, in order to translate these novel therapies to clinical practice, a better understanding of melanoma pathogenesis, its interaction between cancer and immune cells, and the genetic alterations resulting in targetable melanoma-associated antigens (i.e., neoantigens), such as MART-1 (Melan-A), gp100, and tyrosinase (Schumacher and Schreiber 2015; Cervinkova et al. 2017), is required. By utilizing vaccination strategies to increase the number of detectable neoantigens, we may be able to improve recognition of tumor cells by CD8+ T cells, circumventing the immune escape mechanism utilized by the tumor cells (Ragupathi et al. 2000; Terando et al. 2007), increasing tumor cell killing and reducing metastases (Lau et al. 2001; Hepner et al. 2017).

Vaccine development with a BCG adjuvant can induce immune responses contributing to antitumor immunity through the recruitment the NK, CD4+, and CD8+ T cells (Murphy et al. 1993; Kim and Cantor 2014; Kaufmann et al. 2016), resulting in the release of cytokines into the tumor microenvironment (Lardone et al. 2017). BCG-stimulated DCs secretes higher levels of TNF-α, IFN-γ, and IL-12p40 (Lalor et al. 2010; Kumar and Bhaskar 2019).

BCG causes the nonspecific identification of pathogen-associated molecular patterns (PAMPs) by TLR2, TLR4, TLR8, and TLR9, triggering an inflammatory cascade that results in the activation of macrophages (Mantovani and Sica 2010; Iqbal and Hussain 2014; Velmurugan et al. 2016), dendritic cells (Tsuji et al. 2000), and NK cells (Brandau et al. 2001; Kleinnijenhuis et al. 2014). BCG can also increase TRAIL expression and activation of TLR4 and TLR2 in neutrophils, triggering cancer cell apoptosis and improving melanoma patient responses (Ludwig et al. 2004).

It has been shown that the use of BCG as an immunotherapy agent for melanoma can result in tumor regression (Morton et al. 1976) and that topical or intratumoral injection of BCG induces immune cell infiltration and enhances the long-lasting expression of chemokines and cytokines, including CXCL9, CXCL10, CXCL11, IL-15, TNF-α, and IFN-γ (Yang et al. 2017). Due to the high levels of tumor heterogeneity observed clinically and relatively low immunogenicity of melanoma-associated antigens, immunotherapeutic strategies for melanoma utilizing rBCG strains capable of secreting functional cytokines that can stimulate antitumor immune responses show great promise (Sanlorenzo et al. 2014).

The development of recombinant BCG (rBCG) strains has resulted in new vaccines, capable of delivering different antigens that trigger induction of CD8 T cell immune responses with long-lasting memory compared with conventional systems (da Costa et al. 2014; Liang et al. 2015; Oliveira et al. 2017). Also, these rBCG strains have been genetically modified to display different phenotypes, for example through the utilization of Hsp60, Hsp7, pAN, and 18-kDa promoters (Newton-Foot and Gey Van Pittius 2013; Oliveira et al. 2017) that increase cytokine expression and enhance immune responses (Himmelrich et al. 2000; Slobbe et al. 1999). The use of different rBCG strains secreting a variety of cytokines such as IL-4, IL-6, IL-2, IFN-γ, and GM-CSF can modify and potentially improve the antitumor response (Murray et al. 1996; Zhou et al. 2015).

In addition, this method also allows for the development of auxotrophic strains, including strain, where genes involved in the synthesis of metabolites are deleted, strains without antibiotic resistance markers, and strains displaying stable expression both in vitro or in vivo (Borsuk et al. 2007; Seixas et al. 2010; Rizzi et al. 2017).

Several studies have reported the administration of rBCG strains for melanoma treatment (Table 2). One of the most common rBCG strains used for anticancer therapy is rBCG strains expressing interleukin-2 (rBCG-IL-2) and GM-CSF (rBCG-GM-CSF). These combinations increase the production of INF-γ, cytokine important for inhibiting tumor cell growth (Fujimoto et al. 1996).

Table 2 Recombinant BCG developed for melanoma treatment
Full size table

One of the most important strains used is a Pasteur strain, which contains the heat shock protein 60 (hsp60) promoter controlling the expression of IL-2. When administered by intratumoral injection, this rBCG strain displayed immunomodulatory properties and resulted in a 45% reduction in tumor size in a murine B16 melanoma model (C57BL/6 mice) (Duda et al. 1995). In addition, rBCG can express recombinant human interferon-alpha 2B (rhIFN-α) under control of the hsp60 promoter, resulting in increased IFN-γ production and immunostimulatory properties (Luo et al. 2001).

Conclusion

This review summarizes the used of wild-type and various rBCG strains as therapeutic agents for melanoma to induce immunomodulatory activities and enhance antitumor immune responses, as well as its combination with chemotherapy and immunotherapy agents. Few rBCG constructs have been explored and little is known about the immunological basis and mechanisms of action leading to BCG-induced melanoma cell killing. Further investigation of these mechanisms of action will allow for the development of improved treatment strategies to improve the survival of MM patients.