Immunotherapy is a change of paradigm in the treatment of cancer. It focusses the attention in the fight against cancer on the immune system of the cancer patient and tries to translate knowledge from basic immunological research into new strategies of treatment. Such translational research is constantly generating new advances and approaches.

This invited manuscript is a contribution to the Section “State-of-the-Art Biochemistry in Germany”. The corresponding author (VS) is a german biochemist and immunologist, co-author SvG is specialist in paediatric hemato-oncology and co-author WS is specialist in pharmaceutical biology, tumor immunology and translational oncology. The review is authentic with a focus on own achievements embedded in related work and latest findings in the field.

Oncolytic viruses (OVs) are interesting anti-cancer agents with high tumor selectivity. They replicate in and kill cancer cells without damaging healthy cells. This review focusses on the avian OV Newcastle disease virus (NDV) with its inherent anti-neoplastic and immune stimulatory properties. Avian NDV is the most extensively characterized member of the avulaviruses due to the high mortality rate and economic loss caused by virulent strains in the poultry industry. This enveloped paramyxovirus contains a non-segmented, negative-sense, single-stranded RNA genome encoding 6 structural propteins (N, P, L, M, HN, F) involved in the viral life cycle that is limited to the cytoplasm of the host cell. A special property of oncolytic NDV consists of its potential of breaking therapy resistance in human cancer cells. It is therefore particularly suited to counteract the immunosuppressive effects of the tumor microenvironment (TME).

T cell based cancer-specific immune reactivity represents the basis of the types of immunotherapy discussed in this review: post-operative active-specific immunotherapy with NDV-modified vaccines (see 4.11. and 4.15.) and adoptive cellular immunotherapy (see 2.3. to 2.5.). We therefore start with T cells. Chapter 2 presents examples of spontaneous anti-tumor T cell mediated immune responses, as evidenced in particular in the bone marrow (BM). It also provides examples of adoptive T cell mediated immunotherapy studies in which immune memory T cells infiltrate the TME and lead to tumor rejection. Chapter 3 addresses cellular details of the immunosuppressive TME before in chapter 4 the effect of NDV on the various cell types of the TME is being described. While intratumoral application of NDV directly influences the TME, post-operative active-specific immunotherapy with NDV-modified vaccines stimulates tumor-reactive T cells which indirectly affect the TME. The fifth and final chapter reports upon clinical studies of immunotherapy, using glioblastoma multiforme (GBM) with its immunosuppressive TME as an example.

For introduction to the topic, two up-to-date reviews may be helpful, one concerning oncolytic NDV [1], the other cellular immunotherapy [2].

The formation of a solid tumor by a neoplastic cell requires a support ecosystem, i.e., an appropriate TME to allow growth and prevent immune attacks. In absence of a TME, such a transformed cell can principally initiate spontaneous host immune responses via the innate and the adaptive immune system.

This paragraph provides evidence for spontaneous immune T cell reactivity from a well-defined murine tumor model system. It then presents latest insights into spontaneous anti-tumor immune responses in human. Cellular immunotherapy studies involve murine and human immune T cells.

The highly aggressive murine ESb lymphoma, when transplanted into syngeneic mice subcutaneously (sc) or intraperitoneally (ip) grows, metastasizes and kills the host within 2–3 weeks. When transplanted into the ear pinna (ie), however, a site with a high density of dendritic cells (DCs), the cancer cells induce a strong immune response which prevents tumor growth and metastasis [3]. In these immune mice, induced ESb cancer-reactive CD8+ memory T cells (MTCs) control tumor dormancy in the bone marrow (BM) and establish long-term systemic immune resistance upon sc tumor cell challenge [3].

When ESb cells were transfected with the bacterial lacZ gene, it was possible to follow single tumor cells in tissues such as lymph nodes, spleen and BM of ESb-lacZ ie transplanted mice. LacZ, coding for the enzyme ß-galactosidase (Gal), was not only a marker to visualize individual tumor cells but it also served as a surrogate tumor-associated antigen (TAA) that induced major histocompatibility complex (MHC) class I-lacZ-peptide specific CD8 cytotoxic T lymphocytes (CTL) [4].

Research on tumor dormancy in the BM revealed that BM can function as a priming site for spontaneous T cell responses against blood-borne antigens, including TAAs [4]. This was a surprise because textbook immunology teaches that BM is a primary lymphoid organ involved in hemato- and lymphopoiesis while secondary lymphoid organs like lymph nodes and spleen are involved in initiating and facilitating immune responses. Due to tumor-induced angiogenesis, solid tumors become connected to the blood circulatory system so that tumor cells and TAAs can enter the blood. Blood-derived naïve T cells can home to BM sinus endothelium, transmigrate into the parenchyma and interact there with resident CD11c+ DCs. The latter are highly efficient in taking up exogenous blood-borne antigen and processing it via MHC class I and class II (MHC-I and MHC-II) pathways. Upon scanning of the DCs for expression of blood-borne antigens, the T cells with the corresponding T-cell receptor (TCR) form clusters with the antigen-presenting cells (APCs) in BM stroma, become activated, proliferate and differentiate into MTCs. Both, the activated T cells and the MTCs can transmigrate back into the BM sinuses and recirculate via the blood.

A novel tumor model was established for the study of long-term protective immunity and immune T cell memory [5]. Tumor-reactive immune cells against ESb-lacZ tumor cells were generated from a naïve T cell repertoire by a well established ie priming/ip restimulation protocol and transferred to tumor-inoculated T cell deficient nude (nu/nu) mice. The cell transfer prevented tumor outgrowth and resulted in the persistence of a high frequency of Gal-specific CD8+ T cells in the BM and spleen as demonstrated by tetramer staining of CD8 T cells specific for an immunodominant Gal epitope [5]. In contrast, immune cell transfer without tumor cell challenge did not result in detectable levels of Gal-specific CD8 MTCs.

Long-term immune memory and tumor protection could be maintained over four successive transfers of Gal-primed T cells between tumor-inoculated nu/nu recipients. The Gal-specific CD8+ MTCs from the first transfer could be activated and recruited into the peritoneal cavity by ip tumor cell challenge. From there they were harvested for a second adoptive transfer together with tumor cell challenge. About four weeks later the Gal-specific MTCs had returned into a resting state and were detected in the BM. This long-term experiment (>6 months) with four rounds of antigenic restimulation and adoptive immune cell transfer demonstrated longevity and functionality of the MTCs [6]. The results also suggested that the BM microenvironment has special features that are of importance for the maintenance of tumor dormancy and immunological T-cell memory. A low level of persisting Gal antigen appeared to favour the selection of Gal-specific MTCs over irrelevant MTCs in BM niches of CD8+ MTCs [6].

Acquisition of a BM phenotype by recirculating and tissue-resident MTCs has been described [7]. Redirection to the BM of gene-modified T cells was reported to improve T cell persistence and antitumor functions [8]. A dynamic kinetic view of human T cell memory concluded that homeostasis of circulating, proliferating and resting MTCs is controlled by different rheostats: tissue-exit and tissue entry signals for circulating MTCs, proliferation-inducing signals for proliferating MTCs, and availability of a survival niche for tissue-resident, resting MTCs [9].

Primary CD4 and CD8 T-cell responses generated in BM are autonomous and can occur in the absence of classical secondary lymphoid organs (lymph nodes and spleen) [4]. In spite of the absence of molecular adjuvants, the BM T cell responses to blood-borne surrogate TAAs were not tolerogenic and resulted in the generation of CTLs. Primary T-cell responses in BM were also discovered in mice reconstituted with transgenic T cells from OT-I or OT-II mice specific for ovalbumin (OVA) [4]. The BM microenvironment, upon entry of antigens in absence of adjuvants, also facilitates DC (APC):CD4 T cell interactions and maintenance of CD4 memory [10,11]. Why the BM microenvironment does not require adjuvants to initiate primary T cell responses against blood-borne antigens is not yet clear but it is possible that naïve T cells in this environment are in a higher state of activation. An important role for BM as a secondary lymphoid organ was confirmed later by demonstrating high frequencies of Wilms tumor antigen 1 (WT1)-specific CD8+ T cells in BM from tumor-bearing patients [12].

BM represents an excellent site for long-term maintenance of memory CD4+ and CD8+ T cells due to special niches providing survival cytokines such as IL-7 and IL-15 [11,13]. The link between MTCs and stromal cells in survival niches is very robust and can provide efficient memory over a lifetime in tissues such as the BM [14].

Mutation-derived tumor neoantigens play an important role in generating spontaneous anti-tumor immune responses. In recent years molecular pathways have been identified which influence T cell immunoreactivity to tumor neoantigens and cancer immuno-editing. For example, durable CD8+ neoantigen-specific T cell immunity was discovered to be controlled through mRNA m⁶A and the YTHDF1 m⁶A binding protein in DCs involved in cross-priming [15]. Type I interferon (IFN-I) was found to activate APC function in DCs through IFN-stimulated genes (ISG+DCs). Unlike cross-presenting DC1, ISG+DCs acquire and present intact tumor-derived peptide-MHC (pMHC) complexes [16]. In addition to MHC-I neoantigens recognized by CD8+ T cells, MHC-II neoantigens recognized by CD4+ T cells have been identified [17]. These have a key function in shaping tumor immunity and response to immunotherapy.

Spontaneous immune responses to TAAs or tumor neoantigens have been described not only in animal model systems but also in cancer patients. Immune cells can be isolated (i) from tumor samples as tumor-infiltrating lymphocytes (TILs), (ii) from peripheral blood derived mononuclear cells (PBMCs) or (iii) from BM derived mononulear cells.

Spontaneous anti-tumor immune responses in cancer patients could be analyzed from BM aspirate samples. BM samples from 39 primary operated breast cancer patients and 11 healthy females were analyzed for the presence and frequencies of spontaneously induced MTCs with peptide-HLA-A2-restricted reactivity against 10 breast tumor associated TAAs and 3 normal breast tissue-associated antigens in short-term IFN-γ enzyme-linked immunospot (ELISPOT) essays. 67% of the patients recognized TAAs with a mean frequency of 144 TAA reactive cells per 10⁶ T cells. Strong differences of reactivity were noticed between TAAs, ranging from 100% recognition of prostate-specific antigen (p141-149) to only 25% recognition of MUC1 (p12-20) or Her-2/neu (p369-377). Reactivity to normal breast tissue-associated antigens was low [18]. The study revealed the shaping of a polyvalent and highly individual BM T-cell repertoire in cancer patients [19].

Enrichment of MTCs and other profound immunological changes in the BM was reported from untreated breast cancer patients [20]. The proportion of MTCs among CD4+ and CD8+ T cells was much higher in BM from cancer patients than in BM from healthy donors. The extent of MTC increase was related to the size of the primary tumor. Patients with disseminated tumor cells in their BM had more memory CD4+ T cells and more CD56+CD8+ cells than patients with tumor cell-negative BM [20].

BM samples and peripheral blood from 41 pancreatic cancer patients were characterized for location, frequencies and functional potential of spontaneously induced MTCs specific for individual or common TAAs. Pancreatic cancer is highly malignant and dominated by Th2 cytokines in patients` sera suggesting systemic tumor-induced immunosuppression. Surprisingly, high numbers of tumor-reactive T cells were found in all BM samples and in 50% of blood samples. These cells secreted the Th1 cytokine IFN-γ upon stimulation with TAAs [21].

Detailed studies of cognate interactions between MTCs and APCs from the BM of cancer patients revealed bidirectional cell stimulation, survival and antitumor activity in vivo [19]. For example, IFN-α which can be induced in DCs by T cells, has a reciprocal effect on T cells by inducing the expression of IL-12 receptor ß, enabling the T cells to respond to IL-12 and to differentiate into Th1 cells. Other relevant cytokines in this cognate interaction between DCs and CD4+ and CD8+ T cells are IL-2, IFN-γ and TNF-α [22].

Freshly isolated T cells from BM of breast and pancreatic cancer patients recognized autologous tumor cells and rejected them in a xenotransplant model demonstrating their functional and therapeutic potential [23]. In short-term culture with autologous DCs pre-pulsed with tumor lysates, patient`s MTCs from BM (but not from PBMC) could be specifically reactivated to IFN-γ producing and cytotoxic effector cells [23].

A single ip transfer of such restimulated BM T cells into NOD/SCID mice caused regression of autologous tumor xenotransplants. This immune response was associated with infiltration by human T cells and tumor cell apoptosis and necrosis. This demonstrated therapeutic efficiency in vivo of ex vivo re-activated BM-derived cancer-reactive MTCs from cancer patients. Transfer of BM derived CD45RA(-) MTCs but not CD45RA(+) naïve T cells infiltrated autologous tumor but not autologous skin tissue. The TILs had a central or effector memory phenotype and produced perforin. Many of them expressed P-selectin glycoprotein ligand 1 and were found around P-selectin (+) tumor endothelium. Tumor infiltration included cluster formation in tumor tissue by MTCs with co-transferred DCs. Depletion of DCs from restimulation cultures before transfer to NOD/SCID mice reduced therapeutic efficiency suggesting an important contribution of APC restimulation in tumor tissue [24].

These studies demonstrated selective homing of human MTCs to human tumors in xenotransplanted mice and suggested that tumor rejection is based on the recognition of TAAs on tumor cells and DCs by autologous specifically activated central and effector MTCs [23,24].

A review from 2015 described the spontaneous induction of cancer-reactive MTCs from BM, their maintenance by the BM microenvironment and their therapeutic potential [25]. A pilot clinical study investigated adoptive immunotherapy of advanced metastasized breast cancer with BM-derived cancer-reactive MTCs [26]. The BM MTCs apparently had an extensive expansion capacity in the patients [25]. Immunological responder patients showed a significantly longer overall survival (OS) than nonresponders (median survival 58.6 vs. 13.6 months; p = 0.009) [27].

Effective immune rejection of advanced metastasized cancer was demonstrated in a graft-versus-leukemia (GvL) animal model of already cachectic mice [28]. In situ activated tumor-immune T cells, induced in allogeneic, tumor-resistant, MHC identical but superantigen different donor mice could transfer strong GvL effects accompanied by only mild graft-versus-host (GvH) reactivity. A single systemic immune cell transfer into 5 Gy irradiated cachectic DBA/2 mice bearing up to 4 week established syngeneic tumors and macrometastases led to massive infiltration of tumor tissue by CD4+ and CD8+ donor T lymphocytes. Primary tumors of 1.5 cm diameter were encapsulated and rejected from the skin and liver metastases were eradicated. For the first time such adoptive cellular immunotherapy was followed in individual live animals by ³¹P-NMR spectroscopy of primary tumors. This allowed to evaluate changes in tumor tissue pH. An approximately 25,000 fold excess of metastatic tumor cells could be rejected as revealed quantitatively by FACScan analysis of lacZ gene transfected tumor cells [28].

Lessons from such GvL studies in animals about complete remission of cancer in late-stage disease by radiation and transfer of allogeneic MHC-matched immune T cells, in particular of MTCs from the BM [29], were:

(i)

reversion of tumor tissue pH from acid to neutral after 3–4 days as a first sign of the immunotherapeutic effect,

(ii)

donor CD4 T cell infiltration in the tumors 6 days after cell transfer,

(iii)

formation of a broad capsule of fibrous tissue between the tumor area and the skin,

(iv)

tumor rejection and long-term survival,

(v)

wound healing and scar tissue formation at sites of primary tumor rejection (skin) and at sites of metastases (liver and kidney),

(vi)

reconstitution of normal fur at the site of rejected primary tumor,

(vii)

cellular interactions: donor CD4+ and CD8+ immune T-T cell interactions, donor T cell-host macrophage interactions around liver metastases, vß6 donor T cells recognizing a tumor-associated viral superantigen (vSAG-7) interacting with tumor cells and APCs [29],

(viii)

reversibility of a state of cachexia,

(ix)

disproval of the hypothesis that a tumor is a never healing wound [30].

Table 1 lists the most important aspects presented in this chapter.

Table 1. Spontaneous anti-tumor T cell responses.