New clinical advances in immunotherapy for the treatment of solid tumours

Abstract

Advances in understanding the mechanisms of cancer cells for evading the immune system surveillance, including how the immune system modulates the phenotype of tumours, have allowed the development of new therapies that benefit from this complex cellular network to specifically target and destroy cancer cells. Immunotherapy researchers have mainly focused on the discovery of tumour antigens that could confer specificity to immune cells to detect and destroy cancer cells, as well as on the mechanisms leading to an improved activation of effector immune cells. The Food and Drug Administration approval in 2010 of ipilumumab for melanoma treatment and of pembrolizumab in 2014, monoclonal antibodies against T-lymphocyte-associated antigen 4 and programmed cell death 1, respectively, are encouraging examples of how research in this area can successfully translate into clinical use with promising results. Currently, several ongoing clinical trials are in progress testing new anti-cancer therapies based on the enhancement of immune cell activity against tumour antigens. Here we discuss the general concepts related to immunotherapy and the recent application to the treatment of cancer with positive results that support their consideration of clinical application to patients.

Introduction

By the mid-twentieth century, the development of chemotherapy as an adjuvant to radiation therapy and surgery was a significant advance for leukaemia and lymphoma treatment. Later, these approaches impacted the treatment of solid tumours.¹ Nevertheless, chemotherapy, as a systemic unspecific treatment, also affects non-malignant cells and contributes to generalized adverse secondary effects. The search for specific targets to treat cancer developed slowly, but the discovery of specific drugs, such as imatinib by the end of the 1990s, a tyrosine-kinase inhibitor now approved by the The Food and Drug Administration (FDA) for the treatment of several types of cancers, made chemotherapy an efficient targeted therapy. Along these lines, the use of antibodies to target specific proteins on the membrane of tumour cells became especially relevant in cancer treatment and, for the first time in 1997 the FDA approved rituximab for the treatment of B-cell lymphomas. These results led to the subsequent approval of other antibodies, such as trastuzumab and bevacizumab for breast and colon cancer treatment, respectively. Furthermore, the increasing evidence that the cancer microenvironment, composed mainly of extracellular matrix components, fibroblasts, immune cells, among other normal cells, could modulate cancer cell phenotype and proliferation through paracrine and yuxtacrine signalling² led to the notion that these components can control tumour development. Indeed, the fact that immune cells are present as lymphocyte infiltrates in several solid tumours, such as breast, prostate, pancreatic and colon cancer, has revealed the active role of humoral and cellular immune response on tumour suppression.^3–7 Leucocyte infiltrates are composed of cells of lymphoid and myeloid origin, mainly dendritic cells, macrophages, natural killer cells, granulocytes and mast cells, which interact directly and indirectly with tumour and stromal cells to modulate the tumour microenvironment.³ During the first steps of cancer progression, immune cells promote an inflammatory response aimed to repair tissue damage through the secretion of growth factors, interleukins, chemokines, histamine and other bioactive molecules.^2,3,8 Nevertheless, the chronic exposure of tumour cells to these molecules can eventually promote angiogenesis, proliferation and tumour development.² Further, the specific composition of infiltrating immune cells has been correlated with tumour progression, regression and treatment response^5,6,9 and the quantification of infiltrating immune cells can also be used as a prognostic factor, known as ‘immunoscore’.¹⁰

According to the National Cancer Institute definition, immunotherapy is any ‘treatment to boost or restore the ability of the immune system to fight cancer, infections, and other diseases’.¹¹ Under this concept, cytokines are master immunomodulators because they have autocrine and paracrine function, being able to promote or inhibit cellular and humoral responses. In 1998, the use of high-dose interleukin-2 (IL-2) was approved by the FDA to treat melanoma, which was a significant advance in cancer treatment, because by targeting the immune system it was possible to successfully control the disease.¹² However, cytokines alone are not enough to elicit a complete and efficient immune response for cancer treatment, and so far, together with the use of cytokines, monoclonal antibodies and cell-derived vaccines are the most developed approaches for this purpose. Several of these approaches have been evaluated in several clinical trials, confirming high efficiency and specificity in cancer treatment, and in consequence, have been approved by the FDA for cancer treatment.^13–16 Here we resume the basic approaches for immunotherapy development and review some of the most successful clinical trials of the last 4 years for the most common and studied types of solid cancers.

Antigenic targets for immunotherapy

The development of new immunotherapeutic approaches for cancer treatment has developed along with the discovery of new tumour antigens as immunogenic targets.¹⁷ On the one hand, there are tumour antigens that play an essential role in tumour growth or progression and that can be targeted by cytotoxic cells and antibodies.⁸ These antigens are called oncoantigens⁸ and can be divided into three main classes according to their cellular localization. Class I antigens are exposed on the membrane of tumour cell proteins that are over-expressed compared with normal cells, such as the over-expression of human epidermal growth factor receptor 2 (HER2) in breast and gastric cancer, among others. Class II antigens are proteins and other molecules secreted by tumours or environmental molecules, of which certain concentrations facilitate tumour development. Finally, Class III antigens, are intracellular antigens that cannot be recognized by immune cells because they present intracellular localization,¹⁸ especially for those proteins localized in the nucleus.⁸

In addition, there are some tumour-associated antigens (TAAs) that cannot be classified as oncoantigens because they do not necessarily promote tumour development, but rather they can be efficiently used as tumour biomarkers and/or to improve the specific anti-tumour immune reaction. They may be the product of a somatic mutation that causes a truncated protein that is recognized by the antigen-processing machinery and presented on the major histocompatibility complex class I (MHC-I) molecules by tumour cells. Also, oncofetal proteins, normally expressed during fetal development, can be found aberrantly expressed during adulthood, such as the extensively studied α-fetoprotein as well as carcinoembryonic antigen.¹⁹ In the case of secreted antigens, Spondin-2, a secreted extracellular matrix protein has been identified as a reliable diagnostic biomarker for prostate cancer and to predict chemotherapy responses.²⁰ Carcinoembryonic antigen can also be found in serum of other cancer patients, mainly in rectal cancer and non-small-cell lung cancer (NSCLC) patients, is used as a biomarker for chemotherapy response and prognosis.^21–23 Also, we would add exosomes to this class, which are an interesting target for immunotherapy because they express MHC-II and co-stimulatory molecules that can eventually activate tumour antigen-specific T cells. However, it has been also shown in a tumour-bearing mouse model that tumour-derived exosomes can inhibit the anti-tumour immune response.²⁴ Breast cancer and B-cell lymphoma patients can show resistance to monoclonal antibody therapy due the secretion of exosomes containing HER2 and CD20, which sequester trastuzumab and rituximab, respectively, the therapeutic antibodies used for the treatment of these cancers.^25,26 To overcome this problem, a therapeutic haemofiltration approach has been suggested to selectively capture and remove exosomes from the circulatory system.²⁷ In the case of intracellular antigens, new strategies have focused on the liberation of intracellular antigens from the cells to induce a specific immune response. For example, Noguchi et al. showed in a mouse model that combined treatment of chemotherapy plus a targeted monoclonal antibody against Cancer-testis antigen (NY-ESO-1), an intracellular protein expressed by cancer cells but not normal cells (except in testis germ cells²⁸), can induce an efficient CD8⁺ T-cell response.²⁹ Chemotherapy drugs, which induce the release of NY-ESO-1 from necrotic cells, allow anti-NY-ESO-1 antibodies to form in situ immune complexes and induce a higher number of effector/memory NY-ESO-1-specific CD8⁺ T cells that secrete more interferon-γ and/or tumour necrosis factor-α compared with drug treatment alone. These combined therapy-treated mice showed an improved survival due to a long-lasting anti-tumour capacity.²⁹

It is important to point out that most of these antigens are normally expressed by cells in healthy tissues. One of the major challenges for the development of cancer vaccines is the determination of the optimal and most specific tumour antigens to be used as target for immunotherapy. According to this notion, a good antigen should display high antigenicity and a homologous expression in tumour tissue to overcome the escape of tumour cells due to tumour heterogeneity. On the basis of this knowledge, a translational research project carried out by the National Cancer Institute focused on the discovery of new and common tumour antigens among different cancers.³⁰ For this purpose, they selected 75 candidate antigens for further analyses using different approaches to determine their potential use in the development of new anti-cancer vaccines. Even though none of the 75 antigens had all the expected characteristics for a cancer antigen (listed in ref. 30), 46 of them were immunogenic based on clinical trials and 20 were suggested to be potential targets for immunotherapy. Moreover, there are other approaches being used for the discovery of new tumour antigens as potential immunotherapy targets. For example, Walter et al.³¹ used an antigen discovery platform called XPRESIDENT for the development of a new vaccine for the treatment of renal-cell cancer (RCC). This platform allowed the identification, selection and validation of multiple tumour-associated peptides naturally expressed in clinical cases and is based on genomics, proteomics and bioinformatics approaches and on T-cell immunology. Also, ImmunoCore, a biotechnology company located in Abingdon, UK (http://www.immunocore.com/contact/), has focused in the search for specific T-cell targets for the development of cancer therapies. So far, they have created a bi-specific drug called IMCgp100, a soluble, affinity enhanced T-cell receptor specific for a fragment of gp100, a melanoma-specific antigen, fused to an anti-CD3 specific antibody single-chain variable fragment, which confines killer T cells to where melanoma cells are and induces their destruction.³² A phase I/II clinical trial is currently in progress to assess the safety profile and to establish a tolerable dose of IMCgp100 in patients with HLA-A2-positive malignant melanoma (NCT01211262).^33,34

Immunotherapy types

Active therapy comprises the destruction of tumour cells by a direct effect or by indirectly stimulating immune responses.³⁵ One of the most frequently used strategies is to take advantage of soluble molecules, such as cytokines,⁸ which are independent of antigen recognition by host immune cells. Such an approach is known as non-specific-antigen immunotherapy. Interleukin-2 and interferon-α have been approved by the FDA, but there are others soluble molecules, such as IL-7 and granulocyte–macrophage colony-stimulating factor (GM-CSF), that are being used in human trials to enhance anti-cancer therapy (Table1). Also, the allogeneic bone marrow transplant is a recurrent alternative passive therapy for haematological and some solid cancers. Although allogeneic transplant of peripheral blood stem cells has shown better results, it has been observed that the main advantage of bone marrow transplant is a reduced probability of relapsing in some haematological cancers.³⁶

Table 1.

Types of immunotherapy against cancer

	Antigen-specific	Antigen-non-specific
Passive	Adoptive cell transfer: LAK cells, TILs + IL-2 Monoclonal antibodies: Anti-HER2, anti-EGFR, anti-VEGF	Bone marrow transplantation
Active	Cancer vaccines: Tumour antigen-based vaccines Peptide-based vaccines DC-based vaccines Vector-based vaccines Idiotype-based vaccines	Cytokines: IL-2, IL-7, IL-12, IFN-γ, IFN-α, GM-CSF, etc. Immunomodulatory monoclonal antibodies: Anti-PD-1, anti-CTLA-4, anti-CD20, etc.

Abbreviations: DC, dendritic cell; GM-CSF, granulocyte–macrophage colony-stimulating factor; IFN, interferon; IL-2, interleukin-2; LAKs, lymphokine-activated killer; TILs, tumour infiltrating lymphocytes.

Immune molecular checkpoints, which work as main regulators of the tolerance/immunity balance, have become a useful target for immunotherapy across different types of cancer because of the independence of antigen-specific immune reactions. Along these lines, checkpoint receptor inhibitors, such as anti-T-lymphocyte-associated antigen 4 (CTLA-4) and anti-Programmed Cell Death 1 (PD-1) antibodies have demonstrated a high efficacy for the treatment of patients with different types of tumours.³⁷ PD-1 is a co-receptor expressed on peripheral CD4⁺ and CD8⁺ T cells, and the co-stimulatory pathway activated by their ligands, Programmed death-ligand 1 (PD-L1) (B7-H1) and PD-L2 (B7-DC) expressed in peripheral tissues, such as tumour or stromal cells, deliver inhibitory signals that block T-cell proliferation and induce immunological tolerance. PD-L1 is mostly expressed on haematopoietic, non-haematopoietic cells and also tumour cells, in contrast to PD-L2, which is normally restricted to fewer cell types.³⁸ A Phase I clinical trial (NCT00730639) has tested an anti-PD-1 monoclonal antibody called BMS-936558 (also known as MDX-1106 or ONO-4538) in patients with advanced melanoma, RCC, NSCLC, colorectal cancer or castration-resistant prostate cancer.³⁷ Although the disease progression was the most common cause of death, there was more extensive survival in patients with NSCLC, melanoma, or RCC treated with the antibody in comparison to those treated with chemotherapy and/or kinase inhibitors.³⁷ As PD-L1 expression was associated with an objective response to treatment, the authors suggested that its expression might be used as a molecular biomarker to identify those patients that could be treated with anti-PD-1 antibody. Currently, there are two ongoing clinical trials (in phase I and phase II) testing the pharmacodynamic and biological properties of anti-PD-1 antibodies in patients with RCC (NCT01358721 and NCT01354431). In addition to these studies, another phase I clinical trial has been published (NCT00729664), in which an anti-PD-L1 antibody was tested in RCC, NSCLC, melanoma, gastric cancer, ovarian cancer, colorectal cancer, pancreatic cancer and breast cancer.³⁹ An objective response reflected in a durable tumour regression of 6–17% was observed, and a prolonged disease stabilization in patients with metastatic NSCLC, melanoma, RCC or ovarian cancer.³⁹ As NSCLC is considered non-immunogenic and non-responsive to immunotherapies, the 10% of objective response obtained in these patients is remarkably significant and suggests this pathway as a target for therapeutic intervention for this cancer and other related cancers.³⁹ In both studies, neither colon cancer nor pancreatic cancer patients had tumour responses after anti-PD-1 or anti-PD-L1 antibody treatment. The use of anti-PD-1 ligand antibody is a promising approach because inhibitory ligands are found to be expressed in tumour cells in response to the immune system activation, as part of a mechanism called adaptive resistance.⁴⁰ Regarding anti-PD-L1 treatment, two phase I clinical trials (NCT01375842⁴¹ and NCT01375842⁴²) were recently published, in which a monoclonal antibody against PD-L1 (MPDL3280A), was evaluated for safety and tolerability in metastatic urothelial bladder cancer and various cancer types. To test whether PD-L1-positive patients might specifically respond to MPDL3280A, patients were selected based on the expression of PD-L1 in tumour-infiltrating immune cells, which was evaluated by immunohistochemistry. An objective response rate (ORR) equal to 43% was obtained for PD-L1-positive previously treated metastatic urothelial bladder cancer patients after 6 weeks of follow up, 7% of them had a complete response rate. After 12 weeks of follow up, ORR was 52%.⁴¹ Complete and partial response was 18% among all tumour types,⁴² with progression-free survival (PFS) of 18 weeks. Taken together these data suggest that blockade of the interaction between PD-1 and its ligands could significantly contribute to increase the efficiency of tumour immunity induced by other treatments. An extensive review of PD-1 and PD-L1 has been recently published elsewhere.⁴³

Importantly, additional checkpoint molecules, such as T-cell immunoglobulin domain and mucin domain-3 (TIM-3) and lymphocyte activation gene 3 (LAG3), have been found over-expressed in cells^44–46 and their potential use in immunotherapy has been reported in pre-clinical studies.^47–49 Special attention has been paid to LAG-3, which is up-regulated by T cells exhausted by chronic antigen exposure and especially over-expressed in tumour-infiltrating lymphocytes (TILs).^45,46 LAG-3 and PD-1, are co-expressed by TILs,⁴⁵ and display synergistic functions at controlling immunological homeostasis. It has been recently shown that dual inhibition of both receptors decreased tumour growth and enhanced anti-tumour immunity.⁵⁰ Because dual treatment promotes tumour-specific immune responses, with reduced unspecific or self-antigen-specific responses, it can be considered as an interesting strategy for further evaluation in the clinic. Along these lines, a phase I clinical trial is in progress to assess the safety and tolerability of the anti-LAG-3, BMS-986016, alone and in combination with anti-PD-1 in subjects with advanced solid tumours (NCT01968109).

The main aim when designing a new immune approach to treat cancer is not only to specifically target an antigen to direct the immune response against tumour cells only, but also to enhance the natural immune response against it, together with creating immunological memory to obtain a prolonged anti-tumour response. However, this aim becomes difficult because during tumour development, tumour cells might lose antigens for which an immune response was established and also express new tumour-specific antigens because mutations may occur in this process. Hence, antigen-specific immunotherapy requires a well-defined and stable antigen target that can elicit an efficient immune response. Passive, antigen-specific immunotherapies involve the use of the immune system components that are already able to promote an immune response against a specific antigen. Cell-based and antibody-based approaches have been developed for this purpose. The adoptive cell transfer basically comprises the transfer of TILs to patients because there is wide evidence that these immune cells participate in tumour regression in different solid tumours.⁵¹ TILs are obtained from tumours and are cultured with IL-2 to promote the expansion of the tumour antigen-specific T cells. After 6 or 8 weeks of culture, the increased number of cells can be transferred into the host where TILs localize to the tumour and start an in situ anti-tumour reaction.³⁵The use of antibodies generated ex vivo against a particular antigen works also as an adoptive transfer of immunity that generates a specific humoral response. This approach has been widely used for the treatment of diverse types of cancer as will be discussed below. The active specific treatments require host immune-competence and involve the use of vaccines. Several types of vaccines are being developed, from purified antigen vaccines to polyvalent antigen vaccines, depending on the immunotherapy purpose. The main challenge is to generate an immune response to a variety of tumour antigens at the same time. Briefly, the main vaccine subtypes developed so far are (i) tumour antigen-based, (ii) peptide-based, (iii) dendritic cell-based, (iv) vector-based and (v) idiotype-based vaccines (for an exhaustive revision see ref. 35). Hobo et al.,⁵² developed a dendritic cell (DC)-based vaccine with improved immunogenic potential by transfecting DCs with the small-interfering RNAs (siRNAs) of PD-1 ligands combined with a target antigen mRNA electroporation. They could specifically silence PD-L1 and PD-L2 with high efficiency and keep the phenotype and migratory capacity of DCs using a lipid nanoparticle whose cationic lipids efficiently mediate siRNA delivery into the cells. Besides, they electroporated DCs with allogeneic minor histocompatibility antigen mRNA and demonstrated a superior stimulatory potential and an effective ex vivo boost of antigen-specific CD8⁺ effector/memory T-cell responses in transplanted cancer patients. Although T cells over-express PD-1 during T-cell activation, expanded antigen-specific T cells would not be down-modulated because DCs no longer express PD-1 ligands, keeping T cells in a highly activated state. These results allowed authors to suggest that siRNA-modified DCs can be clinically used for boosting the immune response in transplanted cancer patients to induce T-cell-mediated anti-tumour immunity without evoking a response against healthy host cells. Another group has also developed the use of mRNA-based modification of DCs to induce a specific T-cell response against tumour cells whose TAAs have not yet been identified. Yu et al.,⁵³ transfected autologous DCs from osteosarcoma tumour-bearing rats with allogeneic mRNA from osteosarcoma cell lines. Vaccinated rats showed a longer survival due to the induction of a specific long-term T-cell memory immune response.

Because each cancer develops its own unique antigenic profile as the result of the histological origin and the interaction of tumour cells with immune cells and the microenvironment, new anti-cancer strategies developed so far for most common cancer types will be discussed individually in the next sections (Table2). Because it is difficult to review all the immunotherapeutic approaches that are currently being developed for various types of cancer, we will focus on clinical trials that have shown promising results for the most common types of solid tumours.

Table 2.

Selected cancer immunotherapy clinical trials

Clinical trial Reference	Phase	Treatment regimen	Patients	Response	Adverse events	Type
Melanoma
NCT00094653 Hodi et al.13	III	Ipilimumab 3 mg/kg per 3 weeks up to 4 treatments Ipilimumab + gp100 Ipilimumab alone Gp100 alone	676
		HLA*0201 + unresectable stage III or IV melanoma, previously treated with chemotherapy and/or temozolomide with IL-2 doses	OS: 10 and 6·4 month	ipilimumab + gp100 versus gp100 alone, respectively HR: 0·68, P < 0·001 OS: 10·1 ipilimumab alone HR: 0·66, P: 0·003	Grade 3 or 4 irAE in 10–15% of patients with ipilimumab and 3% gp100 alone. 7 deaths related to irAEs	Monoclonal anti-CTLA
NCT00324155 Robert et al.58	III	Ipilimumab 10 mg/kg + DTIC Placebo + DTIC Given at week 1, 4, 7, 10 followed by DTIC alone every 3 weeks until week 22	502 untreated metastatic melanoma stage IIIC or IV patients	OS: 11·2 month ipilimumab + DTIC OS: 9·1 month DTIC + Placebo Higher SR at 1 year (47·3% versus 36·3%) and 2 year (28·5% versus 17·9%) 3 year (20·8% versus 12·2%), respectively HR: 0·72, P < 0·001	Grade 3 or 4 AEs in 56·3% of ipilimumab + DTIC versus 27·5% DTIC + placebo (P < 0·001) No drug related deaths	Monoclonal anti-CTLA
NCT00730639 Topalian et al.68	III	Nivolumab injections every 2 weeks for up to 96 weeks	107 advance melanoma patients	OS: 16·8 month 1 and 2 year SR for nivolumab versus control were 62% and 43%, respectively PFS: 3·7 month 1, 2 year PFS rates: 36% and 27%, respectively Objective responses in 31% of treated patients	Most of AEs within the first 6 months of therapy Grade 3 or 4 pneumonitis (1%), fatigue (32%), rash (23%), diarrhoea (18%) 22% grade 3 or 4 presented treatment-related AEs	Monoclonal Anti-PD-1
NCT01295827 Ribas et al.71	II	Non-randomized cohort of ipilimumab-treated (IPI-T) and ipilimumab-naive (IPI-N) treated with different doses of pembrolizumab A randomized cohort was also treated	411 melanoma IPI-T or IPI-N patients	Durable responses (88% by October, 2013) ORR: 40% IPI-N and 28% in IPI-T PFS: 24 versus 23 weeks, respectively OS not reached (1 year OS: 71%)	12% of patients experienced drug-related grade 3–4 AEs, 4% discontinued due drug-related AEs No drug-related deaths	Monoclonal Anti-PD-1
Hamid et al.70	I	Intravenous pembrolizumab 10 mg/kg every 2 or 3 weeks or 2 mg/kg every 3 weeks	135 patients with advanced melanoma	ORR: 37% across all doses The median PFS: 7 month The estimated median OS was not reached	79% reported drug-related AEs of any grade 3% reported grade 3 or 4 drug-related AEs
Robert et al.16	I	Intravenous pembrolizumab at 2 mg/kg every or 10 mg/kg every 3 weeks	173 patients with advanced melanoma	ORR: 26% at both doses Median PFS: 22 weeks (95% CI 12–36) 2 mg/kg group and 14 weeks (12–24) for the 10 mg/kg group HR 0·84	Most common AEs in any grade were fatigue (33% versus 37%), pruritus (26% versus 19%), and rash (18% versus 18%)
NCT00289016 Senzer et al.76	II	Intratumoral OncoVex injections followed 3 weeks later by injections every 2 weeks	50 stage IIIC and IV melanoma patients	OS > 16 month for all patients. 1 year SR was 58% for all patients and 40% for stage IV M1c patients	Grade 1–2 AEs in 85% of patients. Grade 3 AEs infrequent, 21 severe AE unrelated to treatment	Virus-based vaccine
NCT00243529 Aarntzen et al.83	I/II	Patients were intranodally injected into a tumour-free lymph node	45 stage III and IV melanoma patients	PFS: 34·3 month and OS: not reached for stage III patients PFS: 8·1 versus 2·8 month (P: 0·062) in patients with TAA-specific T cells compared with patients without TAA-specific responses, respectively. OS: 24·1 versus 11 month, respectively, for stage IV patients	No unexpected or severe events reported	Dendritic cell-based vaccine
Breast cancer
NCT00567190 CLEOPATRA Baselga et al.94	III	Pertuzumab + trastuzumab + docetaxel. Placebo + trastuzumab + docetaxel As first-line treatment until disease progression or development of toxic events	808 HER2+ breast cancer patients	PFS = 18·5 versus 12·4 month Pertuzumab versus placebo group, respectively. HR = 0·62, P < 0·001	AEs incidence at any grade was 5% higher in Pertuzumab group than control. Diarrhoea, rash, mucosal inflammation, febrile neutrophenia and dry skin	Monoclonal anti-HER2
NCT008299166 EMILIA Verma et al.96	III	Treatment group received T-DM1 (3·6 mg/kg) intravenously every 21 days Control group self-administered oral lapatinib (1250 mg/day) + oral capecitabine every 12 hr	991 HER2+ breast cancer patients previously treated with trastuzumab plus a taxane	PFS: 9·6 versus 6·4 months in T-DM1 and control group, respectively HR: 0·65, P < 0·001 OS: 30·9 versus 25·1 month, respectively HR: 0·68, P < 0·001 ORR: 43·6% and 30·8%, respectively. P < 0·001	Grade ≥ 3 AEs in 57% versus 41% in treatment and control group, respectively Diarrhoea, nausea, vomiting, palmar-plantar erythrodysaesthesia	Monoclonal anti-Her2 conjugated to a cytotoxic agent
NCT00875979 Miller et al.98	IIa	HER2+ patients treated with T-DM1 (3·6 mg/kg) + pertuzumab (840 mg first time, then 420 mg) once every 3 weeks	64 patients with locally advance or metastatic HER2+ breast cancer	Overall ORR: 41% Overall PFS: 6·6%	The most frequent grade ≥ 3 AEs were thrombocytopenia, fatigue and liver enzyme elevations	Monoclonal anti-Her2 conjugated to a cytotoxic agent
Non-small-cell lung cancer (NSCLC)
NCT00527735 Lynch et al.108	II	Concurrent arm regimen: ipilimumab (10 mg/kg)/paclitaxel/carboplatin followed by placebo/paclitaxel/carboplatin Phased arm regimen: placebo/paclitaxel/carboplatin followed by ipilimumab (10 mg/kg)/paclitaxel/carboplatin Control regimen: placebo/paclitaxel/carboplatin	204 untreated small-cell lung cancer patients	Phased arm, but not concurrent arm, demonstrated an improvement in irPFS compared with chemotherapy (5·7 versus 4·6 months; HR = 0·72; P = 0·05), and OS (12·2 versus 8·3 months)	Overall rates of grade 3/4 irAEs: phased arm (15%), concurrent arm (20%) and control (6%) Most common irAE were rash, pruritus and diarrhoea	Monoclonal anti-CTLA
NCT00527735 Reck et al.109	II		130 untreated extensive-disease-small-cell lung cancer patients.	Phased arm, but not concurrent arm, demonstrastrated an improvement in IRPFS compared with chemotherapy alone (6·4 versus 5·3 months; HR = 0·64; P = 0·03) OS: 12·9 versus 9·9 m	Overall rates of grade 3/4 irAEs: phased arm (17%), concurrent arm (21%) and control (9%). Most common irAEs were rash, pruritus and diarrhoea
NCT00415818 Quoix et al.112	IIb	Patients received up to six cycles of cisplatin–gemcitabine with or without subcutaneous TG4010 each week for 6 weeks	148 advanced NSCLC patients	PFS = 43·2% versus 35%, in TG4010 group and chemotherapy alone group, respectively at 6 month ORR = 41·9% versus 28·4%, in TG4010 group and chemotherapy alone group, respectively, at 6 month	Abdominal pain, injection-site pain, and fever were more common in patients in the TG4010 plus chemotherapy group than the chemotherapy alone group	Antigen associated vaccines
NCT00409188 START Butts et al.113	III	Patients were randomly assigned to tecemotide and placebo groups. Initial therapy of 8 consecutive weekly subcutaneous injections of tecemotide (806 μg lipopeptide) or placebo every 6 weeks until disease progression	1239 unresectable stage III NSCLC patients	No significant difference in OS with the administration of tecemotide after chemoradiotherapy compared with placebo (25·6 versus 22·3 month, respectively) OS = 30·8 versus 20·6 month for tecemotide arm compared with placebo arm, respectively HR = 0·78; P = 0·016	Grade 3–4 AE reported in > 2% frequency with tecemotide were dyspnoea (5% versus 4% in tecemotide versus placebo group), metastases to central nervous system (3% versus 1%), and pneumonia (2% versus 3%)	Antigen-associated vaccine
Prostate cancer
NCT00065442 IMPACT Kantoff et al.15	II	Sipuleucel-T or placebo was intravenously administered every 2 weeks, for a total of 3 infusions	152 metastatic castration-resistant prostate cancer patients	Reduction of 22% of risk of death in treatment versus control group. (HR: 0·78, P = 0·03) Median survival 25·8 versus 21·7 month, respectively	In treatment group, AEs were chills, fatigue, backpain, hypertension, hypokalaemia and muscular weakness. Only one grade 4 AE	Autologous dendritic cell vaccine (mononuclear leucocytes)
NCT00078585 Kantoff et al.117	II	PROSTVAC-VF injections versus control vectors. Priming immunization with rV-PSA-TRICOM with subsequent boosts using rF-PSA-TRICOM, with recombinant GM-CSF	125 castration-resistant metastatic prostate cancer	PFS similar in both groups (P = 0·6) OS: 30% versus 17% in treatment and control group, respectively Median survival: 25·1% versus 16·6%, respectively	Well tolerated. Most AEs were injection site reactions (swelling, pruritus). General disorders also reported: fatigue, chill and nausea	Recombinant vaccinia virus PSA (rV-PSA) vaccine
Renal cell carcinoma (RCC)
NCT00523159 Walter et al.31	I	Vaccinations of IMA901, each preceded by administration of GM-CSF as immunomodulator	28 HLA-A*02+ patients with metastatic RCC	Patients that responded to multiple tumour-associated peptides were more likely to experience stable disease or partial control (P = 0·019) than subjects that responded to 1 tumour-associated peptides or had no response	Well tolerated	Multi-peptide vaccine
	II	Use of single-dose cyclophosphamide (Cy) as additional immunomodulator (+Cy arm) with IMA901 vaccinations	68 HLA-A*02+ patients with metastatic RCC 33 pre-treated (+Cy) 35 not pre-treated (−Cy)	No differences in safety profile in +Cy and −Cy arms PFS = 23·5 versus 14·8 months, respectively (HR: 0·57, P = 0·09) Cy did not alter induction of T cell responses Among immune responders, prolonged survival in +Cy arm. (HR: 0·38, P = 0·04)	2 serious AEs reported: systemic allergic reaction after 12th vaccination (caused by GM-CSF) and grade 3 allergic reaction after 11th vaccination
NCT01354431 Motzer et al.125	II	Nivolumab (0·3, 2, or 10 mg/kg) was administered intravenously every 3 weeks	168 treated patients with metastatic RCC	PFS = 2·7 versus 4 month versus 4·2 month for 0·3-, 2-, and 10-mg/kg groups, respectively ORR = 20%, 22%, and 20% Median OS = 18·2 month, 25·5 month, 24·7 month HR = 0·8, 0·9 in the 2- and 10-mg/kg groups compared with the 0·3-mg/kg group, respectively	73% of patients experienced treatment-related AEs in any grade and 11% experienced a grade 3–4 event Fatigue was the most common treatment-related AE in each group	Monoclonal Anti-PD-1

Abbreviations: AE, adverse event; DTIC, dacarbazine; GM-CSF, granulocyte–macrophage colony-stimulating factor; HR, hazard ratio; irAE, immune-related adverse event; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; TAAs, tumour-associated antigens.

Melanoma

Melanoma accounts for the majority of skin-cancer-related deaths. From 2006 to 2010, incidence rates among white people increased by 2·7% per year, and an estimated 76 100 new cases of melanoma were diagnosed in 2014 in the USA.⁵⁴ Melanoma has been considered as a model tumour for studying anti-tumour immunotherapies because the antigenic profiles are widely characterized for this type of cancer. The first drug treatment against melanoma approved in 1975 by the FDA was dacarbazine (DTIC), an alkyating drug used in chemotherapy that interferes with the proliferation of both cancer and normal cells. In 1998, the use of high-dose IL-2 was approved by the FDA as a treatment for this type of cancer. Nevertheless, although the number of cancer patients treated with IL-2 has rapidly increased since then, there has been a decline in recent years of the use of this therapy because new treatments have been evaluated in recent clinical trials.⁵⁵ Currently, two new agents for melanoma treatment have been successfully developed: vemurafenib (PLX4032), a drug that inhibits the kinase activity of mutated BRAF gene (present in 40–60% of melanoma patients),⁵⁶ which has higher affinity for the V600E mutation than the wild-type protein,⁵⁷ and ipilimumab (MDX-010), a human monoclonal anti-CTLA-4 antibody that blocks T-cell inhibition.

In 2010, Hodi et al.¹³ showed in a phase III trial (NCT00094653), that ipilimumab improved the overall survival (OS) of 676 patients with unresectable Stage III or Stage IV melanoma who were HLA-A*0201 positive. These patients, who previously received chemotherapy treatment (DTIC and/or temozolomide) with IL-2 doses, were randomly assigned in a 3 : 1 : 1 ratio. The first group was treated with ipilimumab plus a gp100 peptide vaccine. The second group was treated with ipilimumab alone and the third one was treated with gp100 alone. Patients treated with ipilimumab either alone or with gp100 showed improved OS (median OS = 10 and 10·1 months, respectively) compared with gq100 alone (median OS = 6·4 months), indicating that ipilumumab could be administered alone, because it seems to be independent of the gp100 effect. The extended OS of previously treated patients correlated with increased CD8⁺ T-cell activation and T regulatory (Treg) cell inhibition. The risk of death of these two groups was reduced by 32% (P < 0·001) and 34% (P < 0·003), respectively, compared with the gp100 control group.¹³ Over 95% of patients in each treatment experiences at least one adverse event (AE) during the trial. The most common immune-related AE (irAE) was tissue inflammation, most frequently in ipilimumab-treated groups. Finally, drug-related deaths occurred in 2·1%, 3·1% and 1·5% in the combined treatment, ipilimumab alone and dp100 alone, respectively. Results of a second phase III trial (NCT00324155) were published by Robert et al.,⁵⁸ in which 502 patients with previously untreated metastatic melanoma stage IIIC or IV were randomly grouped in a 1 : 1 ratio and treated with ipilimumab plus DTIC or DTIC alone. OS was significantly longer in the group receiving ipilimumab plus DTIC (11·2 months) compared with the control group (9·1 months). The risk of disease progression was reduced by 24% (P < 0·0069) and no drug-related deaths occurred in the ipilimumab group. Nevertheless, this treatment also increases proliferation of autoreactive lymphocytes, increasing the risk of autoimmune disease. Several side effects of the use of ipilimumab have been reported, especially autoimmune-related complications, such as colon perforation, and retinopathy.^59–64 These side effects should be carefully followed up in future treatments. Notably, this study supported the hypothesis that the combination of chemotherapy with a reduction of Treg cells is a strategy that can provide a more effective anti-tumour immunity.

The success of ipilimumab in melanoma cancer treatment has motivated the use of other antibodies, such as checkpoint receptor inhibitors. As previously mentioned, PD-1 is a key inhibitory receptor expressed by activated T and B cells. It binds to PD-L1 and PD-L2, expressed on antigen-presenting cells and cancer cells. Phase I clinical trials have shown a good safety profile and durable objective tumour regressions in cancer patients to a PD-1 inhibitor, nivolumab (BMS-936558), including melanoma patients^65,66 even when combined with ipilimumab,⁶⁷ reinforcing the idea that combined therapy against CTLA-4 and PD-1 can produce a better anti-tumour response than monotherapy. In a recent phase III clinical trial (NCT00730639), 107 melanoma patients received intravenous nivolumab every 2 weeks for up to 96 weeks; OS rates of 62% at 1 year and 43% at 2 years were reported, and an improved OS (16·8 months). Objective responses were observed in 31% of patients and 7% had disease stabilization lasting at least 6 months. There were no drug-related deaths. Moreover, after 2 years of follow up for some patients after treatment discontinuation, most AEs occurred within the first 6 months of therapy, and cumulative toxicities were not observed with prolonged drug exposure.⁶⁸ Notably, on 4 September 2014, pembrolizumab (formerly lambrolizumab, MK-3475), was approved by the FDA for treatment of melanoma, being the first anti-PD-1 drug to be approved.⁶⁹ In 2013, it received the ‘Breakthrough Therapy’ designation from the FDA for treating patients with advanced melanoma due to good results in a phase I clinical trial (NCT01295827).⁷⁰ In June 2014, at the 50th Annual Meeting of the American Society of Clinical Oncology (ASCO 2014) in Chicago, preliminary results from the largest study performed so far (KEYNOTE-001, NCT01295827) for this drug reported long-term responses. From 411 patients with ipilimumab-refractory and ipilimumab-naive melanoma, 34% met the Response Evaluation Criteria in Solid Tumors (RECIST) objective response criteria, and by October 2013, 88% had sustained their objective response. Even though the median OS has not been reached, 71% of all patients were alive at 1 year.⁷¹ These results are even more promising because the most common AEs, in any grade (fatigue, pruritus and rash) affected no more than 1–2% of patients, and only 4% of patients discontinued treatment due to a drug-related side effect. Furthermore, patients presented a unique pattern of responses to pembrolizumab, therefore authors proposed an alternative immune-related response criteria to detect cases that do well clinically, but that under RECIST would not be considered as good clinical responses.⁷² Results of a randomized expansion cohort of the phase I KEYNOTE-001 study were published. In this study, 173 patients with unresectable melanoma who were previously treated with at least two doses of ipilimumab and had confirmed disease progression were recruited.¹⁶ Also, patients with BRAF mutations that were treated with approved BRAF and/or MEK inhibitors were included. Patients received either pembrolizumab 2 or 10 mg/kg and were followed for a median time of 8 months. Objective response rate was 26% at both doses (difference 0%, P = 0·96). Median PFS was 22 weeks for the pembrolizumab 2 mg/kg group and 14 weeks for the pembrolizumab 10 mg/kg group [Hazard Ratio (HR) 0·84, 95% CI 0·57–1·23]. Grade 3 or 4 drug-related AEs occurred in only 12% patients and grade 3 or 4 irAE occurred in only three patients. These results reflect a good option for the treatment for ipilimumab-refractory melanoma patients, which do not have other treatment options so far. Also, a phase II study is currently ongoing, the KEYNOTE-002 study (NCT01704287), in which the treatment is being tested on 540 advanced melanoma patients.¹⁶

Allovectin-7, a gene-therapy-based immunotherapy, was designed to boost both adaptive and innate immune responses. It is a bicistronic plasmid/lipid composite that encodes the human leucocyte antigen-B7 heavy-chain gene and the β₂-microglobulin gene. It is locally injected and promotes the allogeneic expression of MHC-I antigen, which is usually lost in melanoma cancer cells.⁷³ In 2010, a phase II clinical trial (NCT00044356) reported that 11·8% of selected stage III/IV metastatic melanoma patients achieved an objective response with median duration of 13·8 months and median OS of 18·8 months.⁷⁴ Nevertheless, two phase III clinical trials (NCT00395070, NCT00003647) failed to reach the primary and secondary efficiency end-points. As well as Allovectin-7, OncoVEX^GM-CSF (Talimogene laherparepvec, T-VEC), is an intralesional therapy that consists of recombinant herpes simplex virus type 1 encoding for GM-CSF. It selectively enters cancer cells⁷⁵ and expresses GM-CSF, inducing cell lysis. Cell debris stimulates the maturation, proliferation and differentiation of dendritic cells, so stimulating a tumour-specific immune response. Results of a phase II clinical trial (NCT00289016) were published in 2009, in which 50 patients, 10 with stage IIIC disease and 40 with stage IV disease (74% of them had previously received dacarbazine/temozolomide or IL-2 pre-treatment) were locally injected and a 26% ORR was obtained. Eight had complete response and five had a partial response; OS was 58% at 1 year and 52% at 24 months.⁷⁶ Regressions of distant lesions were documented. Also, post-treatment phenotypic evaluation of tissue samples revealed that numbers of Treg cells, CD8⁺ T cells and myeloid-derived suppressor cells were higher in the tumour environment than in peripheral blood, as is usually seen for melanoma patients. The accumulation of these suppressor cells was decreased upon treatment with OncoVEX^GM-CSF, and also induced local and systemic MART-1-specific CD8⁺ effector cells.⁷⁷ Due the good results obtained, there is a phase III clinical trial ongoing⁷⁸ (NCT00769704) to evaluate safety and risks of this treatment.

Messenger RNA-transfected DCs have been shown to be the most effective loading strategy to induce immune responses because transfected mRNA can produce multiple and possibly more immunogenic antigenic epitopes than an antigen-loading strategy.⁷⁹ It is also independent of the patient's HLA haplotype, as HLA restriction is the main limitation for use of peptides as antigens for DC priming. Besides, mRNA insertion into DCs is considered to be safe and clinically applicable because it has a short half-life and does not insert into the host genome,⁸⁰ and can be used for expressing co-stimulatory molecules into DCs such as toll-like receptors.⁸¹ Since the first immunotherapy pilot clinical trial performed in 1998 in melanoma patients using DC-based vaccines,⁸² several clinical trials are now ongoing. Aarntzen et al.⁸³ presented a phase I and phase II clinical trial (NCT00243529) in 45 stage III and stage IV melanoma patients who were vaccinated with autologous mature monocyte-derived DCs electroporated with mRNA coding for the TAAs gp100 and tyrosinase protein, and pulsed with keyhole limpet haemocyanin protein. There was an enhancement of tumour-specific CD8⁺ and CD4⁺ T-cell responses and enhanced production of interferon-γ in stage II patients. In stage III patients, the median PFS was 34·3 months and in Stage IV patients, those that presented TAA-specific T cells tended toward an improved PFS (8·1 months) and OS (24·1 months) compared with patients without TAA-specific responses (PFS = 2·8 months; P = 0·062 and OS = 11 months; P = 0·101). These results presented DC-based vaccination as a promising adjuvant treatment of stage III melanoma patients. The same year, Oshita et al. published the results for a phase II clinical trial (number not reported) in Japanese patients, in which monocyte-enriched fractions were pulsed with a cocktail of five melanoma-specific synthetic peptides restricted to HLA-A2 or A24 (MART-1, gp100, tyrosinase, MAGE-A2, and MAGE-A3; and gp100, tyrosinase, MAGE-A1, MAGE-A2 and MAGE-A3, respectively) and keyhole limpet haemocyanin. Overall survival in the vaccinated group was higher (13·6 months) when compared with non-vaccinated patients (7·3 months).⁸⁴

Breast cancer

HER2 is a transmembrane tyrosine kinase receptor with no direct ligand identified. It is associated with activation of the phosphoinositide 3-kinase and AKT pathways, regulating a complex signal transduction cascade that influences cell cycle progression, cell survival, cell proliferation and cell motility.^85,86 Over-expression of the HER2 oncogene, shown to lead malignant transformation of cells, account for 20–30% of all breast cancer cases.⁸⁷ In 1998, trastuzumab (Herceptin), a monoclonal antibody that binds to domain IV of the extracellular domain of the HER2 receptor, was approved by the FDA for treatment of HER2-positive breast cancer patients and nowadays it is the standard treatment for these patients. This was the first monoclonal antibody approved by the FDA for the treatment of solid tumours. Treatment with trastuzumab implies the direct effect of blocking the HER2 constitutive function, and also prevents cleavage of the HER2 extracellular domain,⁸⁸ which is an extra mechanism for HER2 signalling activation.⁸⁵ Also, indirect effects on inflammatory response take place; leading to antibody-dependent cell-mediated cytotoxicity, and so the induction of an adaptive immune response.^89–91 A 6-year follow-up phase III clinical trial (NCCTG N983) demonstrated that concurrent trastuzumab treatment together with chemotherapy has improved disease-free survival and a better outcome.⁹² Also, results suggested that the mechanism of action of trastuzumab also intensifies chemotherapy-induced cytotoxicity.⁹³ New therapeutic monoclonal antibodies have been developed to increase or complement the effect of trastuzumab, such as Pertuzumab, a humanized monoclonal antibody that binds domain II of HER2. This domain is known to dimerize with other HER receptors, especially with HER3, potentiating together the HER2 oncoactivity in tumours over-expressing HER2.⁸⁵ Baselga et al.⁹⁴ published the results of the first phase III clinical trial (CLEOPATRA, NCT00567190) in which trastuzumab and pertuzumab plus docetaxel combined therapy was evaluated in 808 patients with HER2⁺ metastatic breast cancer. They were randomly grouped in a 1 : 1 ratio; one group received trastuzumab plus docetaxel and the other received trastuzumab plus pertuzumab plus docetaxel. The median PFS was improved by 6·1 months, 12·4 months in the control group, as compared with 18·5 months in the pertuzumab group (P < 0·001), and a reduced risk of progression or death (HR = 0·62) was reported. Importantly, even though increased rates of cardiac dysfunction with trastuzumab have been reported in earlier clinical trials, the combination of pertuzumab plus trastuzumab plus docetaxel did not increase the incidence of cardiac AEs compared with the control group. The benefits of combined targeting of HER2 reported in CLEOPATRA were later confirmed after a one additional year follow up.⁹⁵ Despite the success of these therapies for HER2-positive metastatic breast cancer treatment, still many patients eventually progress and or present severe toxic side effects and an effort to develop more efficient and less toxic therapies was necessary. EMILIA, a phase III clinical trial (NCT00829166) evaluated the efficacy and safety of trastuzumab emtansine (T-DM1), a monoclonal anti-HER2 antibody linked to a microtubule-inhibitory agent, mertansine (DM1), in patients with HER2-positive advanced breast cancer who have previously received trastuzumab and one of the taxane chemotherapies.⁹⁶ This methodology allowed delivery of chemotherapy directly to HER2-over-expressing cells. Nine hundred and ninety-one patients were randomly assigned in a 1 : 1 ratio to T-DM1 (n = 495) or lapatinib plus capecitabine (n = 496). Treatment with T-DM1 improved PFS, with a median survival of 9·6 months compared with 6·4 months with lapatinib plus capecitabine, and improved OS of 30·9 versus 25·1 months, respectively. These results allowed T-DM1 to be approved by the FDA the same year for patients with HER2-positive, metastatic breast cancer who have previously received trastuzumab or one of the taxane chemotherapies (paclitaxel or docetaxel). Second end-point studies have supporter the high efficacy and tolerability of T-DM1 for the treatment of women with HER2-positive breast cancer.⁹⁷ Recently, a phase IIa clinical trial (NCT00875979) has reported the safety and efficacy of the combined antibody therapy using T-DM1 plus pertuzumab in HER2-positive locally advanced breast cancer or metastatic breast cancer patients.⁹⁸ They reported an objective response rate of 41% (26 of 64 patients) and a median PFS of 6·6 months, along with an acceptable safety profile. The phase III clinical trial (MARIANNE, NCT01120184) is now evaluating safety and efficacy of T-DM1 plus pertuzumab versus T-DM1 plus placebo versus trastuzumab plus a taxane in patients with HER2-positive, progressive, or recurrent locally advanced or chemotherapy-naive metastatic breast cancer.

In addition to the immunotherapy approaches mentioned above, other groups are focused on the development of vaccines that may improve the immune response against a specific target, providing a less cytotoxic effect on non-malignant cells. HER2 peptides, such as E7 (amino acids 369–377 = KIFGSLAFL, also known as NeuVax) and GP2 (amino acids 654–662 = IISAVVGIL) have shown high immunogenicity and its use for anti-cancer vaccines has been proposed and tested. In a phase II study in 195 patients, after 24 months of E7 vaccine treatment, the treated group (E7 plus GM-CSF) showed a 57% reduction in relative risk of recurrence compared with the control group, suggesting that E7 vaccine treatment is a safe and effective option for stimulation of HER2-specific immunity.⁹⁹ Given that it has been previously described, using tumour cell line-based experiments, that trastuzumab and GP2 might act synergistically to stimulate cytotoxic T cells ex vivo,¹⁰⁰ the authors suggest that GP2 vaccination during trastuzumab treatment might be an efficient complementary anti-cancer treatment. In addition, in 2010, Carmichael et al. presented the first phase I trial in which GP2 vaccine was tested in 18 breast cancer patients to document safety, immunological responses and epitope spreading. The drug was shown to be safe and well tolerated.¹⁰¹ Currently, there is a phase II clinical trial (NCT00524277) evaluating the efficacy of GP2 compared with GM-CSF in treating patients with breast cancer.

Cryoablation is a procedure that has already been successfully tested in breast cancer patients.^102,103 It involves the insertion of a cryoprobe into the tumour monitored by a magnetic resonance imaging system, and the tissues are then frozen in vivo to kill cancer cells. Afterwards, broken cells will be recognized by the patient's immune system.¹⁰⁴ In addition, patients are treated with ipilimumab to keep T cells active and to induce memory T cells to prevent cancer returning in the future. It is a promissory approximation because it has been demonstrated that mice with breast cancer show increased production of tumour-specific T cells, eradication of micrometastasis and a better outcome after cryoablation depending on the freezing methods used.^104,105 There is an ongoing pilot study (NCT01502592) for a small group of pre-operative early stage/resectable breast cancer patients, in which the safety of a combined strategy based on magnetic resonance imaging-guided cryoablation and immunotherapy will be evaluated.

Non-small-cell lung cancer

Lung cancer is one of the leading causes of death by cancer in men and women worldwide. NSCLC comprises approximately 84% of all lung cancer cases.¹⁰⁶ Although NSCLC was thought to be poorly immunogenic, recent clinical trials testing checkpoint inhibitors and cancer vaccines have shown promising results, and these approaches are now in late-phase development. Given the positive results obtained for ipilimumab in melanoma,¹⁰⁷ this drug was tested in a phase II clinical trial (NCT00527735). In 204 previously untreated NSCLC patients it was used together with carboplatin and placlitaxel.¹⁰⁸ The treatment improved ipilimumab efficacy in tumours with squamous histology in patients under a phased ipilimumab regimen when compared with a concurrent ipilimumab regimen. In the phased arm both PFS and OS improved (PFS = 5·7 versus 4·6 months; HR, 0·72; P = 0·05, and OS = 12·2 versus 8·3 months, as compared with paclitaxel and carboplatin alone). In parallel, Reck et al.¹⁰⁹ presented the results for 130 extensive-disease (ED-) NSCLC from the same clinical trial. There was an improvement in PFS in patients treated with the phased ipilimumab regimen, but not in those given concurrent ipilimumab, compared with chemotherapy alone (6·4 versus 5·3 months; HR = 0·64; P = 0·03) and an improved OS (12·9 versus 9·9 months). In both studies, irAEs involved skin (rash and pruritus) and gastrointestinal tract (diarrhoea), occurring more frequently in patients treated with ipilimumab. After these results, two phase III clinical trials are in progress to test the phased regimen with ipilimumab in contrast to standard chemotherapy in NSCLC and ED-NSCLC patients (NCT01285609 and NCT01450761, respectively).

Regarding anti-PD-1 treatment, a phase I clinical trial has demonstrated promissory results. Nivolumab was tested for activity in 122 NSCLC using a dose escalation method of 1·0, 3·0, or 10 mg/kg every 2 weeks in an 8-week cycle. ORR was observed in 33% of patients with squamous histology and 12% of patients with non-squamous histology.³⁷ Two phase III clinical trials comparing nivolumab with docetaxel in patients with squamous and non-squamous NSCLC (NCT01642004 and NCT01673867; respectively) are currently active.

MUC1 is a highly glycosylated transmembrane protein that is expressed on the apical membrane of normal cells¹¹⁰ and aberrantly expressed and glycoslated in tumour cells, exposing tumour-associated epitopes.¹¹¹ TG4010 is an antigen-based vaccine composed of the genetically modified vaccinia virus Ankara containing the MUC1 sequence and IL-2, which acts as adjuvant. Results from a phase IIB clinical trial (NCT00415818) were published by Quoix et al.¹¹² One hundred and forty-eight patients with advanced NSCLC, positive for MUC1 expression by immunohistochemistry, were separated in two groups. Group 1 was treated with chemotherapy in combination with TG4010 and group 2 was treated with chemotherapy alone. After 6 months of treatment, PFS and ORR were higher in the TG4010 plus chemotherapy group (PFS = 43·2%; ORR = 41·9%) compared with the chemotherapy-alone group (PFS = 35%, ORR = 28·4%). There were no significant differences of median OS between both groups; however, median overall survival for patients who presented an objective response was higher in the TG4010 plus chemotherapy group compared with the chemotherapy alone group (23·3 versus 12·5 months). In general, the most common AEs in both groups were anaemia, neutropenia and thrombocytopenia. Abdominal pain, injection-site pain and fever were more common in patients in the TG4010 plus chemotherapy group than the chemotherapy alone group. These results suggest that TG4010 is able to improve first-line treatment in advanced NSCLC. Tecemotide (L-BLP25) is a MUC1-derived 25-amino-acid BLP25 lipopeptide vaccine that was tested in a phase III trial (START, NCT00409188). In this study, 829 and 410 patients were randomly assigned to tecemotide and placebo group, respectively.¹¹³ Although the drug failed to meet the primary end-points regarding OS, where 25·6 months were reported in the tecemotide arm, compared with 22·3 months in the placebo group (HR = 0·88; P = 0·12), interesting results were obtained in a subset of patients receiving initial concurrent chemoradiotherapy. OS was higher for patients that received tecemotide compared with patients of the placebo arm (OS = 30·8 versus 20·6 months; respectively; HR = 0·78; P = 0·016). No statistical differences were observed between tecemotide and placebo for any AE. The clinical benefit obtained in the subgroup of patients that received concurrent chemoradiotherapy was not observed previously in other approaches against NSCLC. The START2 clinical trial will assess the efficacy, safety and tolerability of tecemotide in patients with unresectable advanced NSCLC who have showed a response or stable disease after platinum-based concurrent chemoradiotherapy.

Prostate and renal cell carcinoma

Prostate-specific antigen (PSA) is expressed by prostate cells, and naturally specific T helper type 1 responses to this antigen have been found in patients with metastatic prostate cancer.¹¹⁴ However, the prevalence and magnitude of detectable responses to PSA are low and clinically insignificant.¹¹⁴ Under the concept that the low response to the antigen can be positively modulated, vaccines have been designed to improve the immune recognition of this weak immunogenic antigen and induce a specific T-cell response. Among the cell-based vaccines tested so far, only one has been approved by the FDA for the treatment of asymptomatic and minimally symptomatic metastatic castration-resistant prostate cancer, sipuleucel-T, a DC-based vaccine that demonstrated successful efficacy in phase III clinical trials.^14,115,116 This immunostimulant treatment was catalogued as ‘category 1’ or highly recommended vaccine for the treatment of hormone-refractory prostate cancer. The largest Phase III clinical trial performed is IMPACT¹⁵ (NCT00065442), where 512 patients were assigned in a 2 : 1 ratio. One group (n = 341) received sipuleucel-T and the other (n = 171) received placebo, both administered intravenously every 2 weeks. There was a relative reduction of 22% in the risk of death in the sipuleucel-T group as compared with the placebo group (HR = 0·78). This reduction represented a 4·1-month improvement in median survival (25·8 versus 21·7 months, respectively). There is an ongoing Phase I clinical trial (NCT01832870) designed to quantify the immune response and determine the tolerability of sipuleucel-T when given in combination with ipilimumab for patients with advanced prostate cancer. Also, a phase III clinical (NCT01322490) trial is currently ongoing for a poxviral-based PSA vaccine, PROSTVAC-VF (PSA-TRICOM). Previous results for a phase II clinical trial were published in 2010 (NCT00078585),¹¹⁷ in which 82 patients received PROSTVAC and 40 received control vectors. After 3 years of treatment, OS was higher for PROSTVAC-treated patients (25·1 versus 16·6 months in control group), although the study failed to find an association between PROSTVAC-treated patients and PFS.

Renal-cell carcinoma is the most frequent type of kidney cancer, accounting for 90% of adult cases although the incidence varies with geographic location worldwide.¹¹⁸ Walter et al., have developed a tumour-associated peptide-based vaccine named IMA901 through a discovery platform called XPRESIDENT that allowed the detection of nine naturally expressed tumour-specific antigens from primary renal tumour tissues. The vaccine was tested in HLA-A*02⁺ patients with advanced RCC in a phase II clinical trial (NCT00523159) in which T-cell responses to IMA901 and clinical benefit association were assessed.³¹ The treatment was safe and well tolerated, although there were two cases of serious AEs associated with allergic reactions. There was a prolonged survival in patients treated with IMA901 and pre-treated with cyclophosphamide (OS = 23·5 months) versus non-pre-treated (OS = 14·8 months, HR = 0·57; P = 0·09) and an association was established between immune response and OS because there was a prolonged survival among immune responders pre-treated with cyclophosphamide compared with subjects without this pre-treatment (HR = 0·38; P = 0·040). Finally, MDSC4 and MDSC5, phenotypes for myeloid-derived suppressor cells, were negatively associated with OS, and T-cell receptor-ζ expression tended to associate positively, suggesting them as cellular biomarkers. APOA1 and CCL17 were suggested as serum biomarkers for vaccine-induced immunity and OS. Now, a phase III clinical trial (IMPRINT, NCT01265901) is ongoing to assess whether IMA901 can prolong OS in patients with metastatic and/or locally advanced RCC when given along with sunitinib, a tyrosine kinase inhibitor. This multiantigen-based vaccine is a promissory approach for RCC treatment because other vaccines, such as Reniale, an autologous RCC tumour cell lysate; Vitespen an autologous tumour-derived heat-shock protein Gp96 preparation; and Trovax, a single-tumour antigen vaccine, have not been approved for clinical use because phase III clinical trials have not reached the expected results.¹¹⁹

Several studies have shown that PD-L1 expression in RCC tumours is associated with rapid progression and poor outcomes.^120–122 On the other hand, PD-1 is expressed in TILs from clear-cell RCC and has been described as an indicator of poor prognosis, poor survival and as a predictor of latent distant metastasis of clear-cell RCC.^123,124 Results of a recent phase II trial (NCT01354431) in which nivolumab was tested for activity and safety in previously treated patients with metastatic RCC have been published.¹²⁵ One hundred and sixty-eight patients were grouped in a 1 : 1 : 1 ratio and received nivolumab (0·3, 2, or 10 mg/kg) administered intravenously every 3 weeks. Median PFS was 2·7 months (80%), 4·0 months (80%), and 4·2 months (80%) and ORR was 20% (n = 12), 22% (n = 12), and 20% (n = 11) in the 0·3-, 2-, and 10-mg/kg groups, respectively. The median OS was 18·2, 25·5 and 24·7 months in the 0·3-, 2-, and 10-mg/kg groups, respectively, within a minimum follow up of 24 months. Seventy-three per cent of patients experienced treatment-related AEs in any grade and 11% experienced a grade 3–4 event. Forty per cent of patients had received two or three previous anti-angiogenic drugs and one-third had received everolimus previously. In this regard, two phase III clinical trials are currently evaluating metastatic RCC patients. One of them is focused on comparing nivolumab with everolimus in patients pre-treated with anti-angiogenic therapy and the other is assesing the combination of nivolumab and ipilimumab (NCT01668784 and NCT02231749, respectively).

1. Persisting challenges for cancer immunotherapy due patient heterogeneity

It is evident that the therapies discussed above still fail to improve OS for a high percentage of patients, bringing a small benefit or even only adverse events. The question as to why immunotherapeutic approaches work in some patients and fail in others remains unanswered. Every tumour develops distinct mechanisms to evade the immune system, which could confer special resistance to certain immunotherapy. Efforts are being made to identify the subsets of patients in which immunotherapy could work more efficiently. The main strategies used so far are the detection of prognostic and predictive biomarkers through gene expression profiling,¹²⁶ as well as immune peptidome analyses of HLA-bound peptides and epitope prediction by informatics approaches for TAA identification. Notably, a recent study has suggested why ipilimumab works for some patients and not for others by analysing mutation-derived antigen identification through whole exome sequencing of DNA from tumour cells.¹²⁷ The results showed that tumours of ipilimumab responders have a higher mutational burden than those from patients that did not respond to the therapy. A higher mutational rate gives rise to a higher number of neoantigens that can activate T cells after CTLA-4 blockade. However, not all tumours with a high mutation burden responded to the therapy. These studies were extended by using bioinformatic tools to identify among the responders group a subset of antigens that could elicit an augmented anti-tumour immune response in these patients.¹²⁷ Interestingly, whole exome-sequencing and bioinformatic approaches were used to determine neoantigens in mouse models, leading to the identification of tumour-specific neoantigens that, when used to immunize mice, induce T-cell responses that rejected established tumours.¹²⁸ Immunization with tumour neoantigens together with targeting PD-1 or CTLA-4 can induce T-cell activation and rejection of tumours expressing those neoantigens.¹²⁹ These results highlight the importance of using new methodologies, such as whole exome sequencing and bioinformatic tools to identify neoepitopes for the development of personalized immunotherapy that could improve clinical benefit in non-responder patients.

It should be considered when testing new therapies that not all patients manifest a specific clinical response at the same time and with the same magnitude. The most used response evaluation criteria, RECIST or World Health Organization, which determine patient's response, sometimes fail to define whether a new treatment is really efficient because they do not consider the initial immune response effect on the tumour. For example, an increase in tumour burden followed by a reduction or the appearance of new tumour lesions as a product of the reclusion of immune cells to the tumours would be considered as progressive disease.¹³⁰ To consider these alternative immune-response-related scenarios, a new criterion has been proposed – the immune-related response criteria.¹³⁰ These criteria allow the recognition of optimal responses that otherwise would not be recognized by the two classical response evaluation criteria. In spite of this, many recent works still use RECIST for patient's response evaluation.¹³¹ Moreover, the identification of biomarkers for treatment response and others that could predict a patient's response to a given treatment is an issue that must also be considered to determine the efficiency of treatment and to avoid cytotoxic effects and low response ratios, respectively.

Concluding remarks

Even though our knowledge of the complexity of the immune cell interactions that underlie the complete immune response against tumour antigens is still far from complete, new therapies have successfully demonstrated their ability to down-modulate tumour progression through intervening or improving different steps of the immune response process.

Tumour microenvironment must not be ignored. Different cell types cooperatively establish the tumour immune profile;⁴⁰ this is why immunotherapy must not only look for tumour properties to be targeted but also modulate all features as far as possible. Combined treatments of chemotherapy plus immunotherapy have so far shown a better response and OS in the treatment of many cancer types. However, there is still much to do to improve the management of immune responses because there are plenty of promising therapies focused on the exclusive attack of cancer cells, avoiding the systemic toxicity produced by general drugs. The use of combined treatment of chemotherapy plus immunotherapy has been extensively discussed and studied.^58,94 The principal goal of these combined approaches is not only to alter the internal signalling pathways of cancer cells, but also to target extrinsic cells and mediators of cancer cell proliferation. Combined therapy has shown excellent results improving overall survival and tumour response in clinical trials, becoming the best approach for cancer treatment.

Disclosures

Authors declare no competitive interests.

Glossary

Abbreviations:

AE

adverse event

CTLA-4

T-lymphocyte-associated antigen 4

DC

dendritic cell

DTIC

dacarbazine

FDA

Food and Drug Administration

GM-CSF

granulocyte–macrophage colony-stimulating factor

HER2

human epidermal growth factor receptor 2

HR

hazard ratio

IL-2

interleukin-2

irAE

immune-related AE

LAG-3

lymphocyte activation gene 3

NSCLC

non-small cell lung cancer

NY-ESO-1

cancer-testis antigen

ORR

objective response rate

OS

overall survival

PD-1

programmed cell death 1

PD-L1

programmed death-ligand 1

PFS

progression-free survival

PSA

prostate-specific antigen

RCC

renal-cell cancer

RECIST

Response Evaluation Criteria in Solid Tumours

siRNA

small interfering RNA

TAAs

tumour-associated antigens

T-DM1

trastuzumab emtansine

TILs

tumour-infiltrating lymphocytes

Treg

T regulatory cells