Advances in the field of immunotherapy and targeted therapy have brought about new recommendations for patients with melanoma.
Melanoma, a type of skin cancer that forms from melanocytes (pigment-producing cells), is a complex disease that is mostly curable in the early states but considerably more difficult to treat in the later stages.1,2
In the year 2011, melanomas were rated as the seventh-most common type of cancer, occurring at a rate of 19.7 cases per 100,000 individuals, according to recent data from the Centers for Disease Control and Prevention.3 Although melanomas account for fewer than 2% of skin cancers, melanoma causes the vast majority of skin cancer—related deaths.4 In 2011, a total of 65,647 patients developed new cases of melanoma (38,415 men and 27,232 women), and a total of 9128 deaths were reported (6001 men and 3127 women).5 The American Cancer Society projects that 73,870 new cases of melanoma will be diagnosed in 2015, with the majority occurring in men (42,670 cases) and a smaller proportion occurring in women (31,200 cases).4 In 2015, a total of 9940 deaths are expected to occur, approximately two-thirds of which will occur in men.4 Melanoma rates have been rising for at least the past 3 decades. Between 2002 and 2006, the incidence of melanoma grew 33% in men and 23% in women—faster than any other cancer with the exception of lung cancer.6 Between the intervals of 2002-2006 and 2007-2011, the number of adults in the United States receiving treatment for skin cancer grew from 3.4 million to 4.9 million.7 Furthermore, the average annual cost of treatment between 2002-2006 and 2007-2011 more than doubled, from $3.6 billion per year to $8.1 billion per year.7,8
Although the mean age at diagnosis is 62 years (median 59 years), melanoma is one of the most common cancers in individuals under the age of 30 years, and it is particularly likely to occur at a younger age in patients with a family history of disease.4,6 The lifetime risk of developing melanoma varies among different ethnic populations: it is 2.4% for Caucasians, 0.1% for people of African descent, and 0.5% for Hispanics. Some risk factors for melanoma are modifiable while others are not. Non-modifiable risk factors include age (the risk of developing melanoma rises with age, although many cases occur before the age of 30 years); family history (approximately 10% of cases of melanoma occur in patients with a family history of the disease, suggesting that genetic factors have an important role in treatment); a history of prior melanoma; multiple clinically atypical moles or dysplastic nevi; and immunosuppression or use of immunosuppressive medications, as patients with weakened immune systems, including those taking immunosuppressive medications and patients with the human immunodeficiency virus, are more likely than other individuals to develop melanoma.
Modifiable risk factors include ultraviolet light exposure and frequent sunburns (particularly during childhood). A person with a history of melanoma has an increased risk of developing recurrent melanoma, and approximately 5% of those who have had melanoma will develop it again.4 As a result, continual monitoring is required for anyone who has had a case of melanoma.
It is reported that 2% to 5% of patients have metastatic melanoma at diagnosis. Approximately 82% to 85% of new diagnoses of melanoma are detected when the disease is localized, and another 10% to 13% of patients present with regional disease at first diagnosis.6 A number of predictive factors have been used to determine survival. For example, factors that are predictive of poor survival include advanced age (>70 years) and high levels of lactate dehydrogenase.4
Melanoma may initially present as a suspicious pigmented lesion, with the presence or absence of melanoma confirmed by biopsy. Following the biopsy, a pathology report may be ordered to evaluate and record the thickness of the lesion, whether or not the lesion is ulcerated, the mitotic rate, the status of tumor margins, the presence or absence of microsatellitosis, and other clinical factors. Following this report, patients may be evaluated through a skin exam and assessed for melanomarelated risk factors (eg, family history, history of melanoma or other skin lesions), whether or not the tumor is draining to lymph nodes; additional tests may also be performed as indicated.6 Stage III legions may be treated differently depending on whether or not melanoma has spread to sentinel nodes, or if metastases are in transit.6
Reprinted with permission from Balch CM, Gershenwald JE, Soong SJ, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol. 2009;27(36):6199-6206. © 2009 American Society of Clinical Oncology. All rights reserved.
Melanoma that progresses beyond the local site becomes more aggressive and difficult to manage, with the risk of recurrence increasing steadily from 48% at stage IIIA to 71% at IIIB, 85% at IIIC, and 87% at stage IV.9,10 A diagnosis of melanoma reduces life expectancy by approximately 20 years.6 Furthermore, as the disease progresses, the rates of 5-year survival decrease over time: from 78% in stage IIIA disease to 59% in stage IIIB disease and to 40% in stage IIIC disease.1 As the survival curves (Figure 1)1 for stage III melanoma suggest, the disease characteristics of melanoma can vary widely within substages, a factor that can increase the complexity of patient care.1
Current Approaches to Melanoma Treatment
In many cases, the initial treatment of stage 0 in situ melanoma, stage IA melanomas, and stage IB melanomas is wide surgical excision of the lesion, with continuing follow-up and monitoring. However, in the case of stage IA and stage IB melanomas, depending on the presence or absence of ulceration, the tumor thickness, and the mitotic rate, sentinel lymph node biopsy may be indicated after wide excision of the primary melanoma lesion. Careful monitoring is recommended after initial resection. In patients with stage IB or stage II melanomas with thicker tumors (eg, 0.76-mm to 1.0-mm thick), and depending on the presence of ulceration and the mitotic rate, wide excision with or without a sentinel lymph node biopsy may be recommended. After initial resection, treatment depends on the stage of the tumor. In patients who have stage IB or IIA disease after initial surgical therapy, observation or a clinical trial are treatment options. In patients with stage IIB or IIC disease, interferon (IFN) alpha treatment is an additional option. In some cases, patients will develop stage III disease after initial resection of a stage IB or stage II melanoma lesion.6
Treatment of disease may vary, depending on several factors. In patients with stage III disease, treatment varies based on lymph node status. If pathologic testing indicates melanoma has spread to lymph nodes, complete lymph node dissection is recommended, followed by a clinical trial, observation, or IFN alpha therapy. If pathologic testing does not indicate melanoma, but spread to lymph nodes is suspected clinically, wide excision of the primary tumor in addition to complete lymph node dissection is recommended, followed by a clinical trial, observation, or IFN alpha therapy. When an in-transit stage III melanoma lesion is suspected (characterized by intralymphatic tumor in skin or subcutaneous tissue more than 2 cm from the primary tumor but not beyond the nearest regional lymph node basin), treatment may involve one of several strategies, including (preferably) a clinical trial; local therapy with surgical resection (if feasible); intralesional injection (with Bacillus Calmette—Guérin, IFN, or interleukin-2 [IL-2]); local ablation therapy, topical imiquimod, or radiotherapy (for unresectable disease); regional therapy with isolated limb infusion/perfusion with melphalan; or systemic therapy. For stage IV metastatic disease, treatment may involve aggressive surgical management with or without systemic therapy. For patients with distant unresectable metastases, a clinical trial, systemic therapy, palliative therapy/radiotherapy, and best supportive care are the remaining treatment options.6
In these patients, melanoma cells should be tested for the presence or absence of the BRAF V600 mutation, and the patient’s prognosis should be estimated and recorded, as the treatment choice depends partly on the patient’s prognosis (ie, whether or not the patient is expected to deteriorate clinically within 12 weeks). In patients with wild-type BRAF melanoma who are not expected to clinically deteriorate within 12 weeks, and have not experienced progression after initial systemic therapy, systemic treatment options include pembrolizumab, nivolumab, ipilimumab, and high-dose IL-2. In patients with the BRAF V600 mutation who are expected to clinically deteriorate within 12 weeks, dabrafenib plus trametinib therapy is the preferred treatment, followed by other options such as vemurafenib, dabrafenib, pembrolizumab, or nivolumab. In patients with wild-type BRAF mutation who are expected to clinically deteriorate within 12 weeks, or have experienced progression after initial systemic therapy, systemic treatment options include pembrolizumab, nivolumab, cytotoxic agents (dacarbazine, temozolomide, paclitaxel, albumin-bound paclitaxel, or carboplatin/paclitaxel), imatinib (if the activating C-KIT mutation is present), or biochemotherapy (dacarbazine/ cisplatin/vinblastine ± IL-2 ± IFN alpha, or temozolomide/ cisplatin/vinblastine ± IL-2 ± IFN alpha).6
Alternately, in patients with the BRAF V600 mutation who are not expected to clinically deteriorate within 12 weeks, pembrolizumab, nivolumab, ipilimumab, dabrafenib plus trametinib, and high-dose IL-2 are treatment options. In patients with the BRAF V600 mutation who have received initial systemic treatment progress despite maximal targeted therapy with BRAF inhibitors, if they are still healthy enough to receive treatment (dependent on performance status), other systemic therapy options may be pursued, including pembrolizumab, nivolumab, cytotoxic agents (dacarbazine, temozolomide, paclitaxel, albuminbound paclitaxel, or carboplatin/paclitaxel), imatinib (if the activating C-KIT mutation is present), or biochemotherapy (dacarbazine/cisplatin/vinblastine ± IL-2 ± IFN alpha, or temozolomide/cisplatin/vinblastine ± IL-2 ± IFN alpha).6
The Evolution of Immunotherapy in Melanoma
Because immunotherapy targets the immune system and not simply the tumor, its therapeutic use marks an entirely different way of treating cancer, although it is not necessarily a new modality.11 Immunotherapy as a method of cancer treatment has been attempted for over a century, starting with William Coley, who observed spontaneous remissions in patients with cancer who received a series of killed infectious agents known as Coley’s toxin.12,13 Coley developed the toxin after he observed that patients who developed wounds or infections following cancer surgery had better outcomes than those without infections, prompting the theory of immune system stimulation as cancer therapy.12 Although some significant responses were reported during the subsequent 40 years, interest in cancer immunology later waned as radiation therapy and chemotherapy gained more attention.13 However, in the 1980s and 1990s, interest in cancer immunology was renewed as the Cancer Research Institute was founded to evaluate immunologic advances.14 Over time, the role of the immune system in fighting cancer became better appreciated and characterized.13 One of the earliest immunological therapies studied in the field of melanoma was IL-2, as researchers began to learn that IL-2 could be used to turn the immune system on or off. A retrospective analysis of 270 patients with melanoma treated with high-dose IL-2 showed that as a growth-differentiating factor for T cells, IL-2 cured about 6% of patients with melanoma. However, there were substantial side effects associated with the use of IL-2, and response rates were very low.15,16 In 1998, a high-dose bolus of IL-2 was approved by the US Food and Drug Administration (FDA) for the treatment of metastatic melanoma based on its potential to produce durable complete responses (CRs) in a small number of patients.16 The administration of high-dose IL-2 requires hospitalization with intensive monitoring, and because of the risk of multiple organ failure, it can only be used in younger patients with excellent performance status and organ function.16 Another immunotherapy, IFN alpha, also gained approval for the treatment of melanoma, although it is also associated with low response rates and high-dose toxicity despite exhibiting impressive survival responses in a small subset of patients. It is difficult to determine which patients will benefit from treatment with IFN alpha, a factor that limits its use.13 Based on the potential efficacy of immune-based treatments, the search continued for more effective treatments.
One of the goals of continued research in the field of cancer immunotherapy was to determine what factors could be used to overcome the immunosuppressive effects of tumor cells to evade the immune system. One approach to address this challenge was the development of an immune checkpoint blockade, which has been reported to induce tumor regression in several tumor types. The various immune checkpoints, which dampen the immune system effects on tumor cells, are non-redundant and can inhibit T-cell activation, proliferation, and effector function within the lymph nodes and the tumor microenvironment.17 One such checkpoint inhibitor blocks cytotoxic T-lymphocyte antigen-4 (CTLA-4); CTLA-4 plays a role in inhibiting T-cell activation. In 1987, researchers found that the use of antibodies against CTLA-4 could decrease tumor size in mice.11 Another pathway in checkpoint inhibition is blockage of the programmed-death 1 (PD-1) receptor, which contributes to T-cell exhaustion in the peripheral tissues. Research has shown that targeting the PD-1 molecule expressed on dying T cells was another way to brake T-cell activation.11,17
Based on the discoveries of checkpoint pathways, one of the most important developments in the immunological treatment of advanced melanoma came with the FDA approval of ipilimumab, a monoclonal antibody to CTLA- 4. The phase III trial of ipilimumab became the first study to demonstrate extended survival in patients with metastatic melanoma.18 Two separate studies of ipilimumab as monotherapy or together with dacarbazine have demonstrated advantages in overall survival (OS) compared with controls. For example, in a study of previously treated patients, the median OS improved from 6.5 months in patients treated with ipilimumab to 10.1 months (P = .003) in those who received the glycoprotein100 (gp100) vaccine. Furthermore, in previously untreated patients, the use of ipilimumab plus dacarbazine improved survival to 11.2 months from 9.1 months in those who used dacarbazine alone (P <.0001).6 However, CRs were infrequent using ipilimumab alone, and significant immune-mediated complications have been reported.6,18
The OS benefits from the landmark phase III trial of ipilimumab spurred the development of other immunotherapies, including treatments that inhibit the PD-1 pathways or partner molecules. In 2013, Wolchok and colleagues evaluated the use of ipilimumab in combination with nivolumab, a PD-1 inhibitor, based on the theory that CTLA-4 and PD-1 may play complementary roles in regulating adaptive immunity. The study reported that the combined use of ipilimumab and nivolumab was associated with objective response rates that exceeded those previously generated using either agent alone.17 During a large phase I study of pembrolizumab, the overall response rate (ORR) was 38% and the median duration of response was not reached. During an extension of the study in patients previously treated with ipilimumab, the use of pembrolizumab was associated with an ORR of 26%, with a duration of response that was not reached by the end of the study. Furthermore, during a large phase III study of treatment-naïve patients without the BRAF mutation that compared the use of dacarbazine with that of nivolumab, the use of nivolumab was associated with a 1-year OS rate of 73% versus 42%, a median progression-free survival of 5.1 months versus 2.2 months, and an ORR of 40% versus 14%. Although pembrolizumab and nivolumab may cause immune-mediated adverse reactions, the incidence of grade 3/4 toxicities is less common compared with those associated with ipilimumab. The most common adverse events (AEs) associated with the use of PD-1 inhibitors include fatigue, rash, pruritus, cough, diarrhea, decreased appetite, constipation, and arthralgia. There is a consensus that the anti-PD-1 agents nivolumab and pembrolizumab are less toxic and are associated with higher response rates than ipilimumab.6 Of note, in December 2013, Science magazine voted cancer immunotherapy as the “Breakthrough of the Year.”11
Exogenous Vaccines and Adoptive T-cell Therapy
The use of adoptive T-cell transfer is another major avenue of immunotherapy for patients with melanoma that may, in theory, bypass the daunting task of overcoming immune tolerance.13 Marked and durable clinical responses have been reported in patients with melanoma using this approach, and the results have been described as “breathtaking,” with objective tumor responses up to 70%.13,19-21 The goal of this approach is to improve the immune system response to cancer by removing a patient’s T cells and then reintroducing them following a genetic modification or chemical treatment to enhance their activity. However, several drawbacks are associated with this approach, including safety issues associated with the selection of the target, the paucity of such targets, manufacturing complexities and costs, and a lack of durable response in many patients.13 Furthermore, some patients have experienced lethal virus reactivation and other side effects that reduce quality of life.19 As a result, additional trials of adoptive T-cell transfer techniques are currently under way for patients with melanoma, including a phase II trial of genetically modified T cells that express the receptor for the NY-ESO-1 molecule; a phase II trial of 4-1BB tumor-infiltrating cells in melanoma; a phase II trial of 4-1BB, also known as cluster of differentiation (CD) 137 (which has costimulatory activity for activated T cells and can enhance immune activity to eliminate tumors in mice); a phase I/II trial of T cells engineered with 1 of 2 receptors in melanoma; a phase II trial of cellular adoptive immunotherapy using autologous (“self”) CD8+ antigenspecific T cells and anti-CTLA-4 in melanoma; and a trial of ipilimumab and adoptive cell transfer plus high-dose IL-2 in melanoma.14
The area of cancer vaccines as therapy includes an extensive number of investigational strategies in the field of melanoma designed to elicit an immune-mediated response against tumor-specific or tumor-associated antigens, encouraging the immune system to attack cancer cells bearing these antigens. At its most basic level, there are 2 types of vaccines, which include those that are prophylactic and those that are therapeutic. The use of prophylactic vaccines, such as those that prevent the development of cancers associated with the hepatitis B virus or the human papillomavirus, has shown considerable success.13 In contrast, therapeutic cancer vaccines have been difficult to develop, and results thus far have been disappointing, at least for those vaccines that are designed to introduce preselected cancer antigens to stimulate an immune response.13,22 The vaccine Provenge was approved in 2010 as a treatment for advanced prostate cancer and has spurred continued interest in the development of therapeutic vaccines for other tumor types.22 At this time, several phase II trials of vaccines, given alone or with other therapies, are enrolling patients. A multi-peptide vaccine and an indoleamine 2,3-dioxygenase 1 (IDO1) inhibitor, INCB024360 (made by Incyte Corporation), are being tested in a phase II clinical trial in the setting of advanced melanoma (NCT01961115).
It is postulated that IDO1 inhibition may stop the growth of tumor cells by blocking some of the enzymes needed for cell growth. Another trial (NCT02129075) is evaluating the combined use of CDX-1401 vaccine (ie, DEC-205/NY-ESO-1 fusion protein CDX-1401) and the neoadjuvant-based polyICLC vaccine (ie, melanoma poly-ICLC vaccine) with or without the CDX-301 vaccine in patients with stage IIB to IV melanoma. The CDX-1401 vaccine attaches to a protein manufactured by tumor cells and helps the body recognize the tumor to stimulate the immune system, and the CDX- 301 vaccine is thought to aid in the production of dendritic cells (DCs) to attack tumor cells. The poly-ICLC vaccine also stimulates the immune system and may help the DCs reach maturity to recognize tumor cells. The study is intended to determine whether CDX-1401 and poly-ICLC will be more effective when used with CDX-301. A phase IIb clinical trial in patients with stage IV melanoma is evaluating ipilimumab with or without HyperAcute-Melanoma immunotherapy (a vaccine with a series of irradiated allogenic melanoma cell lines [HAM-1, HAM-2, and HAM-3] produced by inserting a mouse gene into human melanoma in an effort to stimulate the immune cells cancer cells) (NCT02054520). A summary of selected melanoma vaccines under development is shown in Table 1.23
The cancer vaccine gp100 has shown improved clinical activity when combined with high-dose IL-2 compared with the use of IL-2 alone. However, results of its use with ipilimumab were disappointing. During a trial of gp100 used in combination with ipilimumab conducted within a phase III trial of patients with melanoma, although the gp100 vaccine demonstrated improved survival when used in combination with ipilimumab, it had no such effects when used alone, and the efficacy of ipilimumab was not improved by the addition of gp100. The lack of efficacy of the gp100 vaccine has been further observed despite the presence of an associated differentiation antigen in more than 90% of melanoma tumors.18 Other disappointing data came from a large phase III study that evaluated the use of MAGE-A3 in patients with melanoma. Although the vaccine did not significantly extend disease-free survival, the trial was planned to continue in a small subset of patients with a certain gene signature. It is thought that the problem with MAGE and many other cancer vaccines is that they cannot prevent immune reactions that dampen the ability of the immune system to mount a successful attack on tumors.24
6MHP indicates 6 melanoma helper peptide; CD40L, cluster of differentiation 40 ligand; GM-CSF, granulocyte-macrophage colonystimulating factor; gp, glycoprotein.
In contrast to the exogenous vaccines described above (that are produced by introducing preselected antigens into an antigen delivery system targeted to DCs or encoded in viral vectors or administered as peptides or proteins in a suitable carrier), the development of endogenous vaccines involves mobilizing antigens from a patient’s own tumors in situ. Compared with exogenous vaccines, endogenous vaccines possess the distinct advantage of potentially allowing the presentation of dozens, or even hundreds, of tumor mutations to the immune system to induce tumor cell death by allowing endogenous DCs to capture, process, and present tumor-derived antigens.13 Endogenous vaccines are now more commonly referred to as oncolytic immunotherapies, and most are still used only experimentally at this time, but their future in advanced melanoma treatment looks promising.
Pathways in Oncolytic Immunotherapy
What Is Oncolytic Immunotherapy?
Oncolytic immunotherapy (OI) is an innovative area of cancer therapy that involves the intratumoral (or less commonly, intravenous)19 injection of a modified virus referred to as an oncolytic virus (OV) that has the potential to induce tumor lysis through selective replication within targeted cancer cells and also through the activation of T cells for a possible specific, systemic immune response that leaves healthy cells relatively unaffected.19,25-27 Thus, in contrast to exogenous vaccines, OIs have the potential to elicit direct toxic effects on cancer cells while also stimulating the adaptive immunity to provide T-cell—mediated therapy.
Therefore, OI is being designed to induce both local and systemic effects, as treatment combines the local effect of an oncolytic virus with the systemic effect of an antitumor immune response.28,29 The goals of such therapy are to use an engineered virus that selectively replicates within tumor cells for an antitumor effect that is oncolytic (ie, exerts direct cytotoxic activity) and also provides immunotherapy by indirectly inducing a systemic immune response.30 Preclinical models have shown that OVs are effective in treating cancer, and the results reported thus far during clinical studies of patients with cancer are quite promising.31
The Historical Basis of Oncolytic Immunotherapy
It has been known for over a century that viruses can elicit an antitumor response. For example, approximately 100 years ago, women with cervical cancer were reported to experience short-term remission of cancer following the administration of a rabies vaccine.32 Decades later, patients with cancer experienced clinical remission following viral infections.33,34 In addition, the inoculation of patients with cancer with crude viral preparations in an attempt to provide therapeutic benefit began at the beginning of the 20th century.34 In the 1950s, the use of viruses to treat cancer began to take hold when rodent cancer models and tissue culture systems were developed and hundreds of patients with cancer were treated with impure oncolytic virus preparations. The viruses usually did not affect tumor growth because the immune system arrested their growth in most cases, with the exception of instances when the infection took hold and tumors regressed, a phenomenon that was often accompanied by illness and death.35 During the 1970s, a purified mumps virus was used and shown to induce tumor regression or decreased tumor size in patients with cancer.36
FMG indicates fusogenic membrane glycoprotein; GM-CSF, granulocyte-macrophage colony-stimulating factor; HSV-1, herpes simplex virus 1; IFN, interferon; IL, interleukin; MCP, monocyte chemotactic protein; SPI, spleen focus forming virus proviral integration oncogene; TK, thymidine kinase.
Later, the use of recombinant DNA technology was used to provide proof of principle and demonstrate the feasibility of using modified viruses in the treatment of cancer. Genetically engineered mutant herpes simplex virus (HSV) killed glioma cells in vitro and inhibited the growth of gliomas in mice.37 These findings laid the groundwork for the engineering of viruses in an attempt to enhance tumor selectivity and the systemic immune response. Although the potential for use of wild-type viruses or vaccines was observed, several limitations were recognized, including infection of both healthy cells and tumor cells, limited potency in tumor cells, a weakened antitumor immune response, a limited accessibility to the tumor, and the potential for causing human disease.25,38-41 The aforementioned factors limited the potential for wild-type viruses/vaccines to be viable treatment options for cancer; therefore, further investigation was warranted. At this time, engineered oncolytic viral treatments are in development for a variety of tumor types.35
Engineering of Viruses for Possible Use in Oncolytic Immunotherapy
The preferential selectivity of OIs for tumor cells over healthy cells is enhanced by viral engineering that includes genetic deletions in the Adenovirus, the HSV, or the Vaccinia virus; the use of tissue- or tumor-specific promoters in the Adenovirus and the HSV; and gene replacement in the Polio virus.38 Furthermore, the antitumor potency may be improved by the engineering of the viral genome; corresponding strategies used to increase potency have included the deletion of gene encoding for the virulence factor (eg, deletion of ICP34.5 in HSV-1) to enable viruses to replicate in tumor cells but not in healthy cells; the insertion of genes that encode cytotoxic proteins to increase virus-induced cell lysis (eg, expression of adenoviral E3 gene coding for adenoviral death protein); the deletion of apoptosis-inhibiting genes to induce cell killing by viral-induced apoptosis (eg, deletion of gene encoding for apoptosis inhibitor E1B 19-kD protein); and the induction of antitumor immune response to both activate and stimulate tumor-specific cytotoxic T cells and to induce immune response against tumor- specific antigens (eg, expression of cytokine genes and expression of tumor-specific antigens) (Table 2).27,29,38,42-44
A number of preclinical studies have shown the enhancement of a systemic antitumor immune response with viral engineering.19,38,45-47 Studies suggested that viruses could be engineered to express and release cytokines such as IL-4, IL-12, monocyte chemotactic protein -1, and IFN,38 and various cytokines may induce local inflammation, increase the expression of major histocompatibility complex (MHC) molecules, and activate antigen-presenting DCs.19,45 Activating cytokine-secreting tumor cells can stimulate DC expansion and infiltration of tumor cells.45 A mouse model that evaluated the intratumoral effects of injecting gliomas with an IL-4 expressing engineered virus showed that the median survival increased from 18 days in the mice injected with saline to 24 days in the mice injected with the engineered HSV-1.47 Other data suggest that increasing the degradation of the extracellular matrix can modulate viral spreading. For example, insertion of the relaxin gene can potentially increase viral load within tumor cells, induce tumor cell apoptosis, and improve survival in mice treated with an engineered virus.39 Insertion of the hyaluronidase gene can be used to increase viral spreading and increase antitumor efficacy.40
Of all vaccine vectors, HSV has the additional advantage of providing direct cytolytic activity against tumor cells with high levels of oncolysis compared with other vectors and a relatively permissive genome that allows the introduction of foreign genes. Two different strains of HSV1 called BL1 and JS1 were isolated from cold sores and compared with HSV1 17, the most lytic laboratory strain available. JS1 was associated with more superior in vitro killing and was therefore chosen as the basis of newer OVs. The enhanced HSV oncolytic virus was engineered by deleting ICP34.5 to prevent neurovirulant HSV infection of non-tumor cells, inserting granulocyte-macrophage colony-stimulating factor (GM-CSF) to produce GM-CSF cytokine to recruit and stimulate DCs to the tumor site, and finally, deleting ICP47 and inserting US11 under the ICP47 promotor to allow the virus to evade protein kinase R, which normally blocks viral protein synthesis, enhancing oncolysis. ICP47 blocks loading of antigens to the MHC, and deletion enhances antigen presentation.42
Guo and colleagues have reported that there is one particular part of the HSV that is critical for overcoming host cell defenses.48 The researchers reported that the HSV enzyme ribonucleotide reductase could trigger or block necroptosis, a form of cell death, depending on what species the host is infecting. In human cells it can block necropotosis, a finding that suggests using non-natural hosts such as mice to study human viral pathogenesis may provide misleading results.48,49 Removal of ribonucleotide reductase from the HSV gene is the partial reason why a modified HSV will replicate in cancer cells,48 as loss of ribonucleotide reductase decreases the ability of modified HSV to replicate in non-dividing cells, increasing its specificity for tumor cells.26 HSV has been observed to mutate over time and acquire ribonucleotide reductase from a host cell; based on the advantage provided, ribonucleotide reductase remained in HSV and soon it acquired other functions that included the suppression of necroptosis. Necroptosis is a backup mechanism to apoptosis that evolved to kill cells and limit viral spread when apoptosis mechanisms become disabled by evolved viral mechanisms that block its activation.49
Mechanism of Action Overview
A variety of both local and systemic effects are thought to account for the proposed mechanism of action of OIs. Locally, selective viral replication is thought to target tumor cells with oncolytic effects caused by tumor cell rupture. 25,26,50 Systemically, T cells attack distant tumor cells and a tumor-specific immune response is indirectly stimulated.19,29,45 There are a number of key players involved in the mechanism of OIs, including the engineered virus itself, tumor cells, activating cytokines, DCs, tumor-specific antigens, and T cells.19,25,29,45,50
Overall, there are 3 major mechanisms by which OVs confer their benefits during cancer therapy.31 The first mechanism leads to direct oncolysis of tumor cells through the direct infection of cancer and endothelial cells in the tumor tissue. The second mechanism is the indirect effects of necrotic/ apoptotic death of uninfected cells induced via antiangiogenesis and the vasculature targeting of the OVs as shown in both animal models and cancer patients.31,51,52 The last mechanism is the activated innate and adaptive tumorspecific immunity. OVs may provide a number of potential advantages over conventional cancer therapy, including tumor selectivity and the development of immune signaling that initiates a more potent antitumor response and generates a potent adaptive antitumor immunity.31
The Cancer Immunity Cycle: Tumor-Derived Antigens Set the Immune System in Motion
The ability of the immune system to continue to eliminate melanoma is dependent on antigen release, antigen presentation, and T-cell memory.53,54 During antigen release, necrosis rather than apoptosis is needed for the release of danger signals, proinflammatory cytokines, and tumor-derived antigens (TDAs), the catalyst for immune response22,31,55-57; during antigen presentation, robust presentation of TDAs is essential to prime and activate T cells, which can initiate an antitumor response.55,57 T-cell memory may help provide long-term immune protection against melanoma recurrence, as repeated and diverse antigen presentation can expand the breadth of immune memory cells.58,59
APC indicates antigen-presenting cell; CTL, cytotoxic T lymphocyte.
Generation of an effective antitumor immune response involves a series of stepwise events that ultimately form a cyclical response that increases the depth and breadth of the immune response against tumor-associated antigens. In cancer patients, this cycle functions suboptimally, allowing cancer cells to avoid death.
Reprinted with permission from Chen DS, Mellman I. Immunity. 2013;39:1-10.
The cancer immunity cycle (Figure 2)57 involves a series of multistep immune events triggered by immune-mediated tumor cell death.57 During key steps of this cycle, disruptions can take place that lead to tumor escape and progression. 31,55,56,58
The release of TDAs through necrosis is the catalyst to guide the immune system to cancer.58 Necrosis is a form of immunogenic cell death (ICD), or cell death caused by stressors including pathogens or oxidative stress. ICD is known to drive an immune response through the release of TDAs, danger signals, and proinflammatory cytokines.
Necrosis of tumor cells releases a combination of TDAs and stimulatory factors that trigger an immune response specifically against melanoma.31 DCs are activated by danger signals, innate immune effectors, and proinflammatory cytokines, 31,56 and they engulf and process TDAs, which initiates an immune response against melanoma.31,55,56 However, in patients with melanoma, antigens can evolve and escape immunosurveillance. Melanoma can escape immunosurveillance through phenotypic evolution of TDAs, which may impede the ability of the immune system to detect antigens, suppressing T-cell activation and impairing formation of adaptive immune memory.60
Danger signals and proinflammatory cytokines help provoke an immune response to melanoma. Understanding the cancer immunity cycle in melanoma can help guide clinical direction. Collectively referred to as danger signals, damage- associated molecular patterns (DAMPs) and pathogenassociated molecular patterns (PAMPs) are actively emitted from dying necrotic cells and play a beneficial role in provoking an antitumor response due to their interaction with the immune system.31,56 Both DAMPs and PAMPs are thought to provoke an immune response by binding with pattern recognition receptors on DCs, thereby activating them. Another way in which danger signals provoke an immune response is by inducing inflammation, stimulating the production of proinflammatory cytokines (eg, tumor necrosis factor, IL-1, and IL-6) and chemokines, such as IL-8. Proinflammatory cytokines and chemokines can act as immunogenic signals (Table 3).55,56,61,62
Repeated and diverse antigen presentation is critical to inducing an adaptive immune response.58,59 DCs enable immune detection by capturing and processing TDAs, which promote DC maturation. These now highly specialized DCs subsequently present TDAs to T cells through the MHC, priming and activating them.55 Studies indicate that T cells are activated when approximately 8000 T-cell receptors are triggered by an antigen-presenting cell such as a DC.
Further, melanoma antigens evolve over time to escape immune detection. Therefore, frequent and continual TDA presentation through necrosis is vital to priming and activating T cells against an array of melanoma cells.58,59,63 In patients with melanoma, DC function may be inhibited.58 Melanoma can prevent DCs from processing TDAs and priming and activating T cells by secreting immunosuppressive cytokines that inhibit differentiation, maturation, and function of DCs.58,61
MHC molecules continuously move peptides (and sometimes lipids or carbohydrates) from the interior of human cells to the exterior, where these fragments are recognized by T cells. This presentation helps activate cytotoxic T lymphocytes to destroy cells expressing matching surface peptides.61 When DCs capture TDAs, they process them for expression on MHCs, a phenomenon called cross-presentation. Cross-presentation allows the immune system to recognize antigens that are not otherwise presented or those that may not access DCs directly, such as tumor cell antigens, or virus antigens.64
T-Cell Priming, Activation, and Memory Generation
As previously mentioned, repeated and diverse antigen exposure through tumor cell necrosis activates T cells to help provide long-term immune protection against melanoma mutations.58,59 T-cell memory formation begins when naïve T cells are exposed to TDAs by DCs, causing the T-cell population to expand. Under normal conditions, a subset of T cells may survive as a heterogeneous pool of immune memory cells, which may contribute to a more rapid response (than the primary response) to a rechallenge of TDAs.
Repeated TDA exposure to diverse antigens may expand the pool of immune memory cells, increasing the heterogeneity of the memory T-cell population for the possibility of long-term immune protection against melanoma recurrence.58 Melanoma can prevent T-cell activation and inhibit an active immune response.55 Melanoma has the ability to inhibit T-cell response through the release of immunosuppressive cytokines.57 This causes a blockade of costimulatory interactions between DCs and T cells, preventing T-cell activation.55 Immune checkpoint proteins can inhibit the immune response primarily at the level of T-cell development and proliferation, which may limit the formation of immune memory.57
After T-cell priming and activation (Figure 2),57 the antitumor response takes place later in the cycle. Next, the TDAs migrate via the bloodstream to the tumors, and infiltrate the tumor beds. Finally, tumor cell recognition and interaction via the MHC occurs, and then tumor cell destruction.
To evade normal elimination mechanisms, melanoma can impair T-cell migration, prevent tumor cell recognition, and impede immunologic cell death. Once melanoma progresses, it can modulate cytokines, impairing the migration of T cells into tumors. Immunosuppressive factors in the tumor microenvironment (such as PD-1 ligand or transforming growth factor beta) can impede T-cell—mediated killing of tumor cells. Melanoma can also impede the ability of T cells in the tumor bed to recognize tumor cells, by downregulating the MHC.57
Local and Systemic Effects of OI
The goal of OI is a balanced immune response that allows for effective oncolysis. To achieve this balance, OI needs to selectively block or dampen the antiviral immune response, but not at the cost of a reduced systemic antitumor response. While augmentation of the immune system can potentially be used to increase the therapeutic effect of OI by clearing infected tumor cells and bystander tumor cells, blockade of the immune system can potentially reduce the therapeutic effect by rapidly clearing the virus, inhibiting viral replication, and minimizing direct tumor cell lysis.65
Of note, the delivery of OVs can be achieved through multiple modes of administration, which include intratumoral administration and intravenous administration. There are potential advantages and disadvantages to both. The potential advantages of intratumoral administration include the delivery of the virus directly to tumor tissue, minimizing effects to normal tissue, and the potential for reducing systemic adverse effects. The possible disadvantages include a potentially limited systemic immune response and inaccessibility of some tumors/tumor types. By contrast, the potential advantages of intravenous administration include the increased bioavailability of the virus to induce a systemic immune response, the increased ease of administration, and an increased accessibility to most tumors and tumor types; the disadvantages may include a lack of direct delivery to tumors, decreasing the potential for oncolysis, an increased concern for off-target effects, and the potential for systemic clearance prior to reaching target tumor sites.25 To date, the majority of trials performed with OIs have administered OVs via intratumoral injection, whereas a smaller number have examined regional or intravenous delivery.66
CD indicates cluster of differentiation; IFN, interferon; IL, interleukin; MHC, major histocompatibility complex.
In preclinical models of glioma or prostate cancer, viruses engineered for enhanced tumor-selective replication have been associated with reduced tumor volume. In the glioma mouse model, single or double gene deletion in the HSV led to a significant (P = .014) reduction in average tumor diameter over approximately 1 month compared with a control.29 Likewise, in a mouse model of prostate cancer, use of the adenovirus with or without the E3 gene was associated with a decrease in relative tumor volume (%) over 6 weeks compared with a control (which demonstrated an increase).67
Preclinical studies have shown that OVs can be used to infect and multiply in tumor tissue while leaving normal tissue unaffected. In a murine tumor model of colorectal cancer, the invasion of newly synthesized viruses was visualized after a single injection and continued to be observed throughout the first 7 days assayed.68 A similar effect was demonstrated during a phase l study of an OV in patients with metastatic tumors, where biopsies revealed treatment was indeed cancer selective.69 In addition to releasing newly synthesized viruses, preclinical data demonstrate that oncolytic tumor cell death results in the release of an array of tumor antigens.25,28 The secretion of activating cytokines into the local environment upon cell lysis results in the stimulation and maturation of DCs.25,70 Stimulated DCs present tumor antigens on MHC molecules to CD8+ and CD4+ T cells.13,25 Activated CD8+ T cells then attack virusinfected and uninfected tumor cells.50
OVs also increase the number of T cells in tumor tissue. In an ovarian cancer model, treatment with an OV was associated with an increase in CD8+ T cells.71 It is thought that the systemic effects of OI are mediated by cytotoxic T cells. Cytotoxic T cells that recognize different tumor antigens on tumor cells undergo cell expansion and disperse throughout the body to attack tumor cells.27 This systemic response holds the potential to seek and destroy not only large tumors but also micrometastases and individual cells that evade clinical detection.72 In preclinical models of colorectal adenocarcinoma tumors, the systemic immune response resulted in the destruction of distant tumor cells.
During the colorectal tumor study, a modified oncolytic HSV-1 expressing IL-12 was injected into a murine model, which resulted in marked antitumor effects. Over 20 days post inoculation, tumor volume in both non-injected tumors and tumors injected with engineered HSV-1 expressing IL- 12 decreased over time in mice given the OV compared with those given placebo.73 During a phase l study of patients with liver cancer, the safety and maximum tolerated dose of an intratumorally injected poxvirus encoding an activating cytokine (human GM-CSF) was evaluated. The 10- patient study found that non-injected tumors also responded to treatment, durable objective responses were observed in 30% of patients, and there was no regrowth of tumors at responding sites within 4 to 18 months.43
Investigational OI: T-VEC Clinical Trials Experience
Talimogene laherparepvec (T-VEC, formerly known as OncoVex GM-CSF) is an oncolytic HSV-1 that was designed to selectively replicate in tumor tissue and produce GM-CSF to boost/enhance systemic antitumor immune responses following intratumoral injection. The postulated dual mechanism of action of T-VEC comprises a local oncolytic effect achieved by infection and selective replication of the virus in tumor tissue resulting in tumor cell lysis and local release of tumor antigens, and enhancement of a systemic antitumor immune response by expression of GM-CSF in the tumor microenvironment to recruit and activate antigen- presenting cells (eg, DCs). DCs have the capacity to capture antigens and induce proliferative responses and cytokine production in CD4+ and CD8+ T lymphocytes to perpetuate immune responses against cancer cells.74
Several other OVs are also being studied, including the DNA viruses Adenoviridae in bladder cancer, Poxviridae in hepatocellular carcinoma, and Parvoviridae in glioma, and the ribonucleic acid viruses Paramyxoviridae in myeloma, Picornaviridae and Retroviridae in glioma, Rhabdoviridae in hepatocellular carcinoma, and Reoviridae in squamous cell carcinoma of the head and neck.75 Other OVs that express GM-CSF include the genetically engineered vaccinia virus and pexastimogene devacirepvec, a poxvirus being studied in patients with liver cancer.22 In addition to OVs such as T-VEC, which expresses GM-CSF, other OVs under development express cytokines (eg, IL-2, IL-12, or IL-18), chemokines (eg, CCL5) or costimulatory molecules (eg, B7.1 and CD40L), and many other mediators of immune response; data show some encouraging results.19,22
Among all the OIs being tested in clinical trials, T-VEC is the first treatment in its class to meet the primary end point during a phase III trial.76 As described above, T-VEC has been attenuated by replacing several viral genes with the human GM-CSF gene. Briefly, the virus is based on the HSV-1 strain JS1 in which the genes encoding ICP34.5 and ICP47 have been completely deleted. ICP47 blocks antigen presentation to MHC class I and II molecules, and the deletion of ICP47 results in the increased and earlier expression of the HSV US11 gene to promote replication in tumor cells and greatly improve the tumoricidal effectiveness without decreasing tumor selectivity. In addition, the virus contains the coding sequence for human GM-CSF, which stimulates the maturation, proliferation, and differentiation of DCs, but does not amplify the antitumor immune response generated by lysed tumor products. In preclinical studies, JS1/34.5-/47-/GM-CSF enhanced the immune response, increased surface levels of MHC class I molecules, and induced tumor responses. To preserve sensitivity to clinically effective antiviral agents, the gene for thymidine kinase has not been altered.44
T-VEC has a dual mechanism of action that includes both the direct attack of cancer cells and the stimulation of the immune system to help recognize and destroy cancer cells. As a result, the mechanism of action includes both local effects and systemic effects; the virus and the vector it contains have been shown to participate in direct lysis of the tumor cell following intralesional injection, and T-VEC also induces a systemic immune response that is likely enhanced by the local expression of GM-CSF,77 while remaining susceptible to antiherpetic agents.44 Because a key feature of the adaptive immune system is memory, there is a high likelihood that an agent such as T-VEC can produce a durable response.
Early data from phase I and II studies assessed the safety and efficacy of T-VEC in a small number of patients with melanoma. During the phase I dose-ranging study, an earlier strain of T-VEC was evaluated to determine its safety in patients with various tumor types, including refractory subcutaneous or cutaneous metastases from breast cancer, gastrointestinal cancer, head and neck cancer, and melanoma. 78 The results of this trial showed that T-VEC was well tolerated, and antitumor activity was shown by tumor flattening, shrinkage, and necrosis in a variety of tumor types, including melanoma. Effects in tumors that had not been directly injected with the vaccine were also observed.44,78 A total of 26 patients received either a single dose or 3 doses of T-VEC at 1- to 3-week intervals, with doses ranging from 106 to 108 plaque-forming units (PFUs) per mL; the volume of virus injected depended on the size of the tumor. Although no CRs or partial responses (PRs) were observed during the trial, changes suggestive of clinical response were observed with 3 of 26 patients who had clinically stable disease and 6 of 26 patients who exhibited flattening of injected tumors or uninjected tumors nearby. Biopsies performed in 14 of the 19 available patients showed T-cell infiltrates, replicating virus, and tumor necrosis. The most common systemic side effects were low-grade constitutional ones and flu-like symptoms.78
Following the results of the phase I trial, a phase II trial was performed to assess the clinically efficacy of T-VEC in patients with unresectable stage IIIC (n = 10) and stage IV (n = 40) melanoma as determined by overall tumor response rate and survival. This trial evaluated a new clinical strain of HSV-1 (ie, strain JS1), which is more effective at killing tumor cells than the HSV strains previously used.
Eligible patients initially received a total intratumoral injection of up to 4 mL at 106 PFU/mL to induce seroconversion in seronegative patients. These lower initial doses were used instead of the higher doses from the phase 1 study in an effort to avoid the flu-like reactions observed in higherdose subgroups. Three weeks later, the maximum total dose per visit, defined as up to 4 mL at 108 PFU/mL, was repeated every 2 weeks using ultrasound guidance if necessary. Up to 10 tumors were treated at each visit, with more drug being injected in larger tumors and less drug injected into smaller ones. Treatment could continue up to a maximum of 24 injections if there was evidence of activity following 8 doses, defined as tumor inflammatory reactions or stable disease or better. The primary objective of the study was to evaluate ORR, which was modified to allow biopsy for disease status and to allow for clinically insignificant limited disease progression based on the possibility of delayed, immune-mediated antitumor effects. The median number of injection sets was 6 (mean, 9 sets), and 5 patients received the full course of 24 injection sets.44
Results of the phase II study suggested that T-VEC was effective in patients with unresectable advanced melanoma and that the toxicity profile was limited. The objective response rate (defined as a CR or PR) was 26% (13/50), which included 16% (8/50) of patients obtaining a CR and 10% (5/50) a PR, with the onset of response occurring from 2 to 10 months following the first treatment. Furthermore, 4% (2/50) of patients were disease-free after surgical resection (when eligibility for resection developed following viral treatment). Responses were observed in both injected and distant metastases, with distant responses documented at uninjected sites in the lung, liver, pancreas, regional and distant lymph nodes, and other soft tissue sites. It is interesting to note that 6 patients in the trial experienced disease progression prior to a CR (n = 4) or PR (n = 2), which may reflect a latency period prior to development of the immune response and the timing of tumor growth in relation to timing of the immune response. One-year survival rates were 58% in all patients and 93% in patients who achieved a PR, CR, or surgical CR (a total of 15/50 patients)]. Of particular interest was the durability of responses, which were ongoing from 16 to 40 months following the first dose of T-VEC, despite 74% of all patients in this trial having failed 1 or more prior nonsurgical therapies for active disease. Eight percent of patients (4/50) experienced grade 3 fatigue, 8% (4/50) had grade 3 dyspnea/abnormal breath sounds, and 6% (3/50) had grade 3 asthenia/muscular weakness; all other grade 3 AEs occurred in less than 5% of patients. No grade 4 AEs were reported.44 The high frequency and durability of overall objective response rates together with the promising 1-year and overall survival rates in patients from this phase II trial spurred the evaluation of T-VEC in a larger trial called OPTiM (Oncovex [GM-CSF] Pivotal Trial in Melanoma), a phase 3 trial in 436 patients with injectable, unresectable stage IIIB to stage IV melanoma.79
OPTiM was the first large trial to demonstrate the efficacy of an OI. The results demonstrated a significant (P <.0001) improvement in durable response rate among patients with unresectable stage IIIB to stage IV melanoma who received T-VEC, suggesting that T-VEC may be a novel approach to treating patients with melanoma who have distant or regional metastases.77
Eligible patients had melanoma that was not surgically resectable; stage 3B/C (with or without in-transit disease) or stage IV disease with limited visceral burden (lactase dehydrogenase ≤1.5 times the upper limit of normal, ≤3 visceral metastases [lung lesions excepted] and no lesion >3 cm; any liver lesion must have been stable for >1 month, and brain lesions must have been treated and stable for >2 months); injectable disease, defined as at least 1 cutaneous, subcutaneous, or nodal lesion; measurable disease (lesions or aggregation of lesions ≥10 mm in greatest diameter); and an Eastern Cooperative Oncology Group performance status of 0 to 1. Eligible patients were enrolled from May 2009 to July 2011 and randomized in a 2:1 fashion to receive: (1) T-VEC administered intralesionally with a starting dose of 4 mL or less containing 4×106 PFU/mL, followed 3 weeks later by 4 mL or less containing 4×108 PFU/mL administered every 2 weeks (n = 295); or (2) 125 micrograms per square meter of body surface area of GM-CSF administered subcutaneously daily for 14 days of every 28-day cycle (n = 141). Randomization was stratified according to disease substage, prior use of systemic treatment, site of disease at first recurrence, and the presence of liver metastases, and patients were to remain on treatment for at least 24 weeks despite progression unless the investigator decided to begin a new therapy or therapy became intolerable. The primary outcome was durable response rate, which was defined as a CR or PR lasting at least 6 months and starting within 12 months of treatment. Key secondary outcomes included OS, ORRs, time to treatment failure (defined as the time from the first dose of study treatment until death or development of clinically significant progression for which no objective response was subsequently achieved), and safety. Responses were determined by an independent blinded end point assessment committee, using modified World Health Organization (WHO) criteria based on evaluation of all lesions.
For the OS secondary end point, 290 events were required for the primary analysis to provide 90% power to detect a hazard ratio (HR) of 0.67 with a 2-sided alpha of 0.05. The primary analysis for durable response rate was based on a data cut-off date of December 21, 2012, while the data cutoff date for OS was March 31, 2014. A final analysis of OS is planned after a minimum of 3 years follow-up.79 The median duration of treatment was 10 weeks in the GM-CSF group and 23 weeks in the T-VEC group.77,79 Responses were observed in both injected and distant metastases.79 With T-VEC, the ORR was 26.4% (10.8% of patients experienced a CR and 15.6% achieved a PR), compared with 5.7% in those who received GM-CSF (0.7% of patients had a CR and 5% achieved a PR) (P <.0001 for ORR of T-VEC vs GM-CSF).77,79 The durable response rate with T-VEC was 33% versus 0% with GM-CSF (stage IIIB/C); 16% with T-VEC versus 2% with GM-CSF (stage M1a); 3% with T-VEC versus 4% with GM-CSF (stage M1b); and 8% with T-VEC versus 3% with GM-CSF (stage M1c).77 Overall, the durable response rate was 16.3% in the T-VEC group and 2.1% in the GM-CSF group (P <.0001).77,79 Of note, the trial also showed borderline OS benefits, as interim results found the median OS was 23.3 months in the T-VEC group and 18.9 months in the GM-CSF group (P = .051), with differences in OS becoming more pronounced over time (a 4.6 percentage point difference between groups at 1 year, an 8.5 percentage point difference at 3 years, and an 11.3 percentage point difference at 4 years). An exploratory subgroup analysis by disease stage suggested that patients with stage IIIB/C or stage IV M1a disease may experience more significant OS benefits compared with those who have more advanced disease (ie, stage IV M1a/b disease). In a similar fashion, other exploratory analyses suggested that patients who receive T-VEC as first-line treatment may experience greater OS benefits compared with patients who receive T-VEC as second-line or higher therapy.79
Serious AEs occurred in 26% of patients receiving T-VEC versus 13% of patients receiving GM-CSF,77 and 3.7% of patients discontinued treatment due to treatment-related AEs.80 The most common AEs associated with the use of TVEC were flu-like effects and fatigue, chills, and pyrexia.77 The only grade 3 or 4 AEs occurring in more than 2% of patients was cellulitis, which occurred in 2.1% of patients receiving T-VEC and less than 1% of patients receiving GMCSF. Vitiligo was reported in 5% of patients treated with TVEC and 1% of those who received GM-CSF. Furthermore, there were 10 deaths in the T-VEC group, including 8 caused by progressive disease and 2 due to sepsis or myocardial infection. In the GM-CSF group, there were 2 deaths due to dyspnea or disease progression.77,79
In summary, the OPTiM phase III trial showed that T-VEC significantly (P <.0001) improved the primary end point of durable response rate (16.3%) compared with GM-CSF (2.1%). Furthermore, the ORR was significantly (P <.0001) higher for T-VEC (26.4%) versus GM-CSF (5.7%). Importantly, responses were observed in both injected and uninjected lesions, which included visceral lesions. For OS, there was an improvement that closely approached statistical significance (HR, 0.79; 95% CI, 0.62-1.00; P = .051), and the median OS was increased by 4.4 months using T-VEC. In an exploratory subset analyses, both durable response rate and OS appeared to be more pronounced among patients with no prior non-adjuvant systemic therapy and among patients with stage IIIB to stage IV M1a disease versus patients with more advanced M1b/c disease.79
An extension study of OPTiM was open to patients if they received the maximum treatment allowed in the OPTiM study (ie, 18 months) and did not have clinically relevant disease progression associated with decreased performance status or had a complete response to medication in the OPTiM trial and developed new lesions within 12 months of the end of treatment. Patients were excluded from the extension trial if they experienced grade 3 or 4 treatment-related toxicities during the core trial (except for grade 3 or 4 injection site reactions, fever, or vomiting), grade 3 fatigue lasting over 1 week during the core trial, grade 3 arthralgia or myalgia during the OPTiM study, or grade 2 or higher autoimmune reactions, allergic/urticarial reactions, or other nonhematologic toxicities that required a dose delay or treatment discontinuation. Patients were eligible if they did not receive additional antitumor therapies for melanoma at the end of treatment in OPTiM; however, palliative radiotherapy or surgical resection of melanoma lesions was allowed. The primary objective of the extension phase of the OPTiM phase III trial was to evaluate the safety of extended treatment with T-VEC or GM-CSF in eligible patients, and the secondary objectives were to evaluate objective tumor response rate and durable response rates.
Tumor response rates were evaluated by investigators using modified WHO criteria for tumor response assessment with baseline defined at the beginning of the parent study. Durable response rate was defined as a PR or CR that began at any point within 12 months of treatment initiation and lasted continuously for 6 months or longer, a time frame that included the parent study period.74
A total of 31 patients were enrolled in the extension study (T-VEC, n = 28; GM-CSF, n = 3). Clinical characteristics of patients enrolled in the extension study were consistent with the parent study. The median treatment duration with T-VEC was 1.5 times longer in the extension study than it was in the OPTiM study. In both studies, most reported AEs were grade 1 or 2 in severity and generally did not lead to discontinuation of treatment. The types and grades of AEs in the extension study were similar to those observed in OPTiM, but the frequency of most AEs was less, suggesting that some of the most common AEs may reduce in frequency over time in the T-VEC group. In the extension study, the most common treatment-emergent AEs in the T-VEC group were fatigue (10.7%), chills (17.9%), and pyrexia (17.9%), while the most common treatment-emergent AEs in the GM-CSF group were fatigue (66.7%), nausea (33.3%), chills (33.3%), pyrexia (33.3%), and influenza-like illness (33.3%). No fatal treatment-related AEs were reported in either study. Continued treatment with T-VEC, but not GM-CSF, was associated with improved ORRs compared with those observed in OPTiM. Among patients receiving T-VEC in the extension trial (n = 28), 25% of patients (7/28) went on to have complete resolution of disease and 86% of patients (24/28) maintained stable disease or better.
Continued treatment with T-VEC may benefit patients who complete up to 18 months of treatment with T-VEC without clinically relevant disease progression. Evidence of efficacy without additional toxicity was shown following extended treatment.74
Based on these phase III data, the FDA is considering approving T-VEC as a treatment for patients with regionally or distantly metastatic melanoma. The advisory committee recommended approval in April 2015, and a final decision is expected by October 2015.81
Experimental Use of T-VEC in Combination or as Neoadjuvant Therapy Prior to Surgery
A hallmark of oncology therapy is the combined use of regimens with distinct mechanisms of action to elicit complementary or additive therapeutic effects without overlapping or additive toxicity profiles. Furthermore, activation of a patient’s immune system during a time when concomitant treatments are working to produce tumor reduction and remission may be an ideal strategy to produce long-term durable benefits.13 Therefore, it follows that the study of a novel therapy such as T-VEC is of interest in combination with newer targeted therapies or even as a part of neoadjuvant treatment prior to surgery. Harrington and colleagues reported on the use of T-VEC combined with radiotherapy and cisplatin during a small phase I/II study of 17 patients with untreated stage III/IV squamous cell cancer of the head and neck. After a median follow-up of 29 months, locoregional control of the cancer was 100%, disease-specific survival was 82.4%, and OS was 70.6%.82 A number of ongoing clinical trials are evaluating T-VEC in combination with other agents, such as checkpoint inhibitors, or as neoadjuvant treatment prior to surgery in patients with advanced melanoma. A phase Ib/II multicenter, open-label trial is currently under way to evaluate the safety and efficacy of T-VEC with or without ipilimumab in approximately 200 patients with advanced and unresected melanoma.
Similarly, a phase Ib/II trial is seeking to enroll 110 patients with previously untreated, unresectable stage IIIB to stage IV M1c melanoma to evaluate the use of T-VEC with or without pembrolizumab (ie, MK-3475). A phase II, multicenter, randomized open-label trial is aiming to enroll 150 patients with completely resectable stage IIIB, IIIC, or IV M1a melanoma to estimate the efficacy of T-VEC (6 doses) as neoadjuvant treatment followed by surgery compared with surgery alone; the primary outcome is 2-year recurrence- free survival. An expanded access study is evaluating the use of T-VEC in patients with stage IIIB to stage IV M1c unresectable melanoma, including those with a performance status of 2, who are otherwise ineligible for or could not access comparable alternative studies. This phase IIIb, multicenter, open-label study is being conducted in select European countries. Studies are also evaluating the use of biomarkers to assess efficacy. For example, a single-arm phase 2 trial is currently under way to evaluate the role of the immune response to T-VEC in unresected melanoma.
The target enrollment for the trial is 110 patients with stage IIIB to stage IV M1c melanoma, and the primary outcome variable has been defined as the correlation between baseline intratumoral CD8+ cell density and objective response rates, evaluated over 2 years.23
Advances in the field of immunotherapy and targeted therapy have brought about new recommendations for patients with melanoma. The new National Comprehensive Cancer Network (NCCN) guidelines for the use of systemic therapies in patients with stage III or IV metastatic melanoma now include the use of immunotherapies and targeted treatment as initial treatment options. For example, pembrolizumab, nivolumab, and, for patients with the BRAF V600 mutation, the BRAF inhibitors (eg, vemurafenib, dabrafenib) are now options for systemic therapy management of metastatic melanoma. Furthermore, the immunotherapy checkpoint inhibitors pembrolizumab and nivolumab are now considered first-line treatment options for metastatic melanoma, whereas they were previously considered appropriate for use only after initial treatment with ipilimumab.
To facilitate the use of targeted therapies when appropriate, the newest NCCN guidelines recommend mutational analysis of melanomas in patients receiving routine treatment and those entering clinical trials. However, patients with no evidence of disease should not receive mutational testing.6
Despite the availability of new therapies, the treatment of patients with melanoma remains difficult because tumor cells can evade immune signaling through several mechanisms, including increasing the production of immunosuppressive cytokines (eg, IL-4, IL-6, IL10), the stimulation of immunosuppressive cells such as T regulatory cells and macrophages, and cell signaling disruption mechanisms such as MHC loss in tumor cells, degradation of the T-cell receptor delta chain, and generation of indoleamine 2,3-dioxygenase.83 Other challenges include the potential for cancer to evolve and develop new survival pathways.84 Therefore, additional modalities are needed to further modulate the immune system response to tumors via a variety of pathways.
Continuing to harness the immune system using patient immunity to fight disease will, it is hoped, achieve better outcomes for patients with advanced disease. Future treatment options place a growing emphasis on immunotherapies and targeted therapies, as ongoing clinical studies continue to evaluate the mechanism of enhancing the body’s own immune response to cancer. Immunotherapies are being used in various combinations with different regimens as scientists work to identify the optimal treatment for patients with melanoma. OVs have been designed to selectively replicate in targeted tumor cells and not in normal healthy cells, to induce targeted tumor cell lysis and enhanced oncolytic cell death.85 The induction of local cytokine expression attracts and activates DCs and T cells for a systemic immune response to target tumor cells and induce a systemic immune response that can result in lysis of distant tumor cells.25
Several OIs are currently being investigated in a number of tumor types.70 The development of appropriate pharmacodynamic biomarkers will help to determine if a particular immunotherapy is producing the desired effects and may be useful to identify which patients are most appropriate for which strategies. Furthermore, because the mechanism of immunotherapy is so distinct from that of conventional cytotoxic treatments, which elicit more immediate responses, new metrics for the evaluation of immunotherapy efficacy may be needed.13
T-VEC is the first OI to demonstrate therapeutic benefit against melanoma in a well-controlled, randomized phase III trial. The use of T-VEC improved durable response rate, defined as a PR or CR lasting continuously for at least 6 months and beginning within the first 12 months following treatment, from 2% to 16% compared with GM-CSF, and median OS was improved by 4.4 months.74 The systemic effect of T-VEC was demonstrated by responses in uninjected lesions. Exploratory analyses suggest that there is a particular benefit in patients with limited visceral disease or when T-VEC is given as a first-line therapy. Monotherapy may provide a novel approach to metastatic melanoma. The use of combination treatments that include T-VEC is rational, and ongoing and future treatments will further explore these combinations.79