Medical School · Year 2 · Immunology · includes a quiz and discussion video

Lecture 10: Tumor Immunology

Unit 2.7: Immunology

Learning Objectives

By the end of this lecture, students will be able to:

1. Describe tumor antigens and immune recognition of cancer
2. Explain the mechanisms of immune surveillance and tumor escape
3. Describe the tumor microenvironment and immunosuppression
4. Explain checkpoint inhibitor immunotherapy
5. Describe adoptive cell therapies (CAR-T, TIL)
6. Explain cancer vaccines and other immunotherapies

Lecture Outline

I. Immune Surveillance of Cancer

The concept of cancer immunosurveillance, first proposed by Burnet and Thomas in the 1950s, posits that the immune system continually monitors tissues for malignant transformation and eliminates nascent tumor cells before they can establish clinically apparent disease. This hypothesis has evolved into the more comprehensive framework of cancer immunoediting, which describes three phases of interaction between the immune system and developing tumors. The elimination phase corresponds to classical immunosurveillance, where immune cells successfully recognize and destroy transformed cells, preventing tumor development. The equilibrium phase represents a prolonged stalemate where the immune system contains but does not eradicate tumor cells, potentially lasting years while tumor variants arise under selective pressure. The escape phase occurs when tumor cells that have acquired mechanisms to evade or suppress immunity finally break free of immune control and progress to clinically detectable cancer.

Multiple lines of evidence support the existence of immune surveillance against cancer. Patients with primary or acquired immunodeficiency, including those with AIDS, organ transplant recipients on immunosuppression, and individuals with rare genetic immunodeficiencies, develop cancers at dramatically increased rates, particularly virally-associated malignancies and lymphomas. Spontaneous regression of established tumors, while rare, has been documented in melanoma, renal cell carcinoma, and other cancers, implying that immune mechanisms can sometimes control or eliminate even advanced disease. The presence of tumor-infiltrating lymphocytes (TILs) in many solid tumors correlates with improved prognosis, suggesting that ongoing immune responses limit tumor growth. The dramatic clinical responses to checkpoint inhibitor immunotherapy provide the most compelling contemporary evidence that the immune system has intrinsic capacity to recognize and eliminate cancer when inhibitory signals are blocked.

The immune system employs multiple cell types and effector mechanisms to recognize and eliminate tumor cells. Cytotoxic CD8+ T lymphocytes (CTLs) are considered the primary effectors of anti-tumor immunity, recognizing tumor antigens presented on MHC class I molecules and killing target cells through perforin-granzyme pathways and Fas-FasL interactions. CD4+ Th1 cells provide essential help for CTL responses and activate macrophages through IFN-gamma secretion, promoting M1 polarization toward an anti-tumor phenotype. Natural killer cells can kill tumor cells that have downregulated MHC class I molecules to escape CTL recognition, as reduced MHC class I expression releases NK cells from inhibitory signals while activating receptors such as NKG2D recognize stress ligands often expressed by tumor cells. Macrophages can phagocytose tumor cells and contribute to antigen presentation, while dendritic cells are critical for initiating tumor-specific T cell responses by cross-presenting tumor antigens to naive T cells.

The effector mechanisms employed by immune cells to kill tumor cells parallel those used against infected cells. The perforin-granzyme pathway, utilized by both CTLs and NK cells, involves perforin-mediated pore formation in target cell membranes, allowing entry of granzyme serine proteases that activate caspases and induce apoptosis. IFN-gamma produced by Th1 cells and CTLs has direct anti-proliferative effects on some tumor cells and upregulates MHC expression, enhancing tumor visibility to the adaptive immune system. Antibody-dependent cellular cytotoxicity (ADCC) occurs when NK cells recognize antibody-coated tumor cells through Fc receptors, releasing cytotoxic granules upon engagement. Macrophages can phagocytose antibody-opsonized tumor cells and secrete cytotoxic factors including TNF-alpha, reactive oxygen species, and nitric oxide. Understanding these effector mechanisms informs the design of immunotherapies that aim to enhance natural anti-tumor immunity or redirect immune cells against cancer.

<image> Panel A: Cancer immunoediting phases diagram showing elimination (immune cells destroying transformed cells with lightning bolt symbols), equilibrium (balance scale with immune cells on one side and tumor cells on other, with clock indicating years), and escape (tumor cells overwhelming diminished immune response, growing into clinical cancer), with arrows indicating progression or regression between phases. Panel B: Evidence for immunosurveillance infographic showing increased cancer rates in immunodeficient populations (AIDS patients, transplant recipients, genetic immunodeficiency syndromes) with relative risk statistics, spontaneous tumor regression examples (melanoma, RCC), TIL-prognosis correlation graphs, and checkpoint inhibitor response rates as proof of concept. Panel C: Anti-tumor immune cells and their roles showing CD8+ CTL killing tumor cell via perforin-granzyme with MHC I-peptide recognition, CD4+ Th1 cell providing help and IFN-gamma, NK cell killing MHC I-low tumor via NKG2D activating receptor, dendritic cell cross-presenting tumor antigen to T cell, and M1 macrophage with phagocytic and cytotoxic functions. Panel D: Tumor cell killing mechanisms illustration showing perforin pore formation with granzyme entry inducing apoptotic cascade, Fas-FasL death receptor pathway, IFN-gamma direct anti-proliferative effects, ADCC with NK cell recognizing antibody-coated tumor cell via CD16, and macrophage phagocytosis and cytotoxic mediator release (TNF, ROS, NO). </image>

II. Tumor Antigens

Tumor antigens are molecules expressed by cancer cells that can be recognized by the adaptive immune system, serving as targets for T cells and antibodies. These antigens fall into several categories with differing immunogenicity and therapeutic implications. Tumor-specific antigens (TSAs) are truly unique to tumor cells and not found on any normal tissues, making them ideal targets that can be attacked without concern for autoimmune damage. Tumor-associated antigens (TAAs) are expressed on tumor cells but also present at lower levels on some normal tissues, meaning immune targeting may cause on-target, off-tumor toxicity. Differentiation antigens are normally expressed only on the cell type from which the tumor arose, such as melanocyte antigens in melanoma. Understanding the antigen landscape of individual tumors has become essential for selecting appropriate immunotherapy strategies and predicting responses.

Neoantigens represent the most immunogenic category of tumor antigens and have emerged as a central focus in cancer immunotherapy. These antigens arise from somatic mutations in tumor cell genomes, creating novel peptide sequences not present in any normal tissue and therefore not subject to central tolerance. Point mutations creating amino acid substitutions, frameshift mutations generating novel reading frames, and gene fusions producing chimeric proteins can all generate neoantigens. The immunogenicity of neoantigens depends on their presentation by MHC molecules and recognition by the T cell repertoire; not all mutations generate immunogenic neoantigens, and computational prediction of neoantigen immunogenicity remains imperfect. Tumor mutational burden (TMB), the total number of somatic mutations in a tumor genome, correlates with neoantigen load and predicts response to checkpoint inhibitor immunotherapy. Tumors with high TMB, such as melanoma (due to UV-induced mutations) and microsatellite-unstable colorectal cancer (due to DNA repair defects), tend to respond better to immunotherapy than low-TMB tumors.

Viral antigens represent another category of highly immunogenic tumor-specific antigens in virally-associated cancers. Human papillomavirus (HPV) drives the majority of cervical cancers and an increasing proportion of oropharyngeal cancers; the viral oncoproteins E6 and E7 are required for malignant transformation and are expressed in all HPV-positive tumor cells, making them attractive therapeutic targets. Epstein-Barr virus (EBV) is associated with several malignancies including Hodgkin lymphoma, Burkitt lymphoma, nasopharyngeal carcinoma, and post-transplant lymphoproliferative disorder; viral proteins including EBNA and LMP are potential targets. Hepatitis B and C viruses drive hepatocellular carcinoma, while human T-lymphotropic virus type 1 (HTLV-1) causes adult T-cell leukemia/lymphoma. These viral antigens are foreign to the immune system, not subject to tolerance, and shared among all patients with the same virally-associated cancer, facilitating development of broadly applicable immunotherapies.

Overexpressed antigens are normal cellular proteins present at much higher levels in tumor cells compared to normal tissues, creating a quantitative rather than qualitative distinction. HER2/neu is overexpressed in approximately 20% of breast cancers and some gastric cancers, serving as both a prognostic marker and a therapeutic target for antibodies (trastuzumab) and antibody-drug conjugates. The epidermal growth factor receptor (EGFR) is overexpressed in many epithelial cancers. Cancer-testis antigens such as NY-ESO-1 and the MAGE family are normally expressed only in germ cells but become aberrantly expressed in various cancers; because germ cells lack MHC class I expression, these antigens are effectively tumor-specific for immune targeting. Oncofetal antigens like carcinoembryonic antigen (CEA) and alpha-fetoprotein (AFP) are normally expressed during fetal development but silenced in adult tissues, then reactivated in certain cancers. While less immunogenic than neoantigens, overexpressed antigens can serve as targets for cellular and antibody therapies.

<image> Panel A: Tumor antigen classification diagram showing hierarchy from most tumor-specific (neoantigens, viral antigens) to less specific (overexpressed antigens, differentiation antigens, cancer-testis antigens), with examples and immunogenicity ratings for each category, and indication of on-target off-tumor toxicity risk. Panel B: Neoantigen generation illustration showing normal gene sequence compared to mutated sequence with point mutation, translation producing novel peptide epitope, processing and presentation on MHC class I, and T cell recognition; alongside TMB spectrum showing high-TMB tumors (melanoma, MSI-high CRC) versus low-TMB tumors with correlation to checkpoint inhibitor response rates. Panel C: Viral antigens in cancer showing HPV E6/E7 proteins in cervical cancer cells (required for maintaining transformed phenotype), EBV EBNA and LMP in lymphoma, with notation that these are truly foreign antigens not subject to tolerance and shared among patients, facilitating universal vaccine development. Panel D: Overexpressed antigen examples showing HER2 receptor highly expressed on breast cancer cell membrane compared to normal levels on other epithelium (target for trastuzumab), cancer-testis antigen NY-ESO-1 normally only in germ cells (MHC-negative) but aberrantly expressed in melanoma and other cancers, and CEA oncofetal antigen pattern. </image>

III. Tumor Immune Evasion

Tumor cells employ multiple strategies to evade immune recognition and destruction, explaining why cancers progress despite the presence of tumor antigens and tumor-reactive lymphocytes. Antigen loss or downregulation represents one of the most direct escape mechanisms. Tumors may lose expression of immunodominant antigens through deletion of the genes encoding them, epigenetic silencing, or outgrowth of antigen-negative subclones under immune selection pressure. This process, termed immunoediting, explains why tumors that develop in immunocompetent hosts often lack the same antigens as tumors arising in immunodeficient settings. MHC class I downregulation prevents antigen presentation to CD8+ T cells; mutations in beta-2-microglobulin (essential for MHC class I surface expression), loss of heterozygosity at HLA loci, and epigenetic silencing of MHC genes all occur in tumors. This mechanism allows escape from CTLs but potentially increases vulnerability to NK cells, which are inhibited by MHC class I.

Immune checkpoint molecules, originally evolved to prevent autoimmunity by attenuating T cell responses, are co-opted by tumors to suppress anti-tumor immunity. PD-1 (programmed death-1) is expressed on activated T cells and delivers inhibitory signals upon binding its ligands PD-L1 or PD-L2. Many tumors upregulate PD-L1 expression, either constitutively through oncogenic signaling or adaptively in response to IFN-gamma produced by infiltrating lymphocytes. When T cells encounter PD-L1-expressing tumor cells, the PD-1/PD-L1 interaction triggers T cell exhaustion characterized by reduced proliferation, cytokine production, and cytotoxicity. CTLA-4 (cytotoxic T-lymphocyte-associated protein 4) competes with the costimulatory receptor CD28 for binding to B7 ligands (CD80/CD86) on antigen-presenting cells, transmitting inhibitory signals and removing B7 ligands through transendocytosis. Additional checkpoint molecules including LAG-3, TIM-3, and TIGIT contribute to T cell exhaustion and are targets for next-generation immunotherapies.

Tumors produce immunosuppressive factors that create a local environment hostile to immune effector function. Transforming growth factor-beta (TGF-beta) suppresses T cell proliferation and function, promotes regulatory T cell differentiation, and inhibits NK cell activity. Interleukin-10 inhibits dendritic cell maturation and antigen presentation while suppressing Th1 responses. Vascular endothelial growth factor (VEGF), primarily known for promoting angiogenesis, also inhibits dendritic cell maturation and creates an abnormal vasculature that impedes T cell infiltration. Indoleamine 2,3-dioxygenase (IDO) catabolizes tryptophan to kynurenine, depleting this essential amino acid from the microenvironment and generating immunosuppressive metabolites. Adenosine, generated by tumor cell ectoenzymes CD39 and CD73 from ATP released by dying cells, suppresses effector T cells through A2A receptor signaling. These factors work in concert to create a profoundly immunosuppressive milieu within and around tumors.

Metabolic alterations in the tumor microenvironment further compromise immune function. Rapidly proliferating tumor cells consume enormous quantities of glucose, creating a hypoglycemic environment that starves T cells of their preferred fuel source. The Warburg effect, whereby tumor cells produce lactate even in the presence of oxygen, acidifies the microenvironment; low pH impairs T cell activation, proliferation, and effector function. Tumor hypoxia, resulting from outgrowth of tumor blood supply, stabilizes hypoxia-inducible factors (HIFs) that promote immunosuppressive pathways and adenosine production. Competition for other nutrients including amino acids (arginine, glutamine, tryptophan) further handicaps immune cells. These metabolic challenges represent a distinct axis of immune evasion that complements antigen loss, checkpoint upregulation, and immunosuppressive factor production, and suggest that metabolic reprogramming of immune cells may enhance immunotherapy efficacy.

<image> Panel A: Antigen loss and MHC downregulation mechanisms showing tumor cell evolution under immune pressure: original tumor with strong antigen expression and MHC class I, then immunoedited tumor lacking antigen (gene deletion, epigenetic silencing) or with beta-2-microglobulin mutation preventing MHC I surface expression, allowing CTL escape but creating NK cell vulnerability. Panel B: Checkpoint molecule upregulation diagram showing tumor cell expressing PD-L1 (either constitutively via oncogenic signaling or induced by IFN-gamma), engaging PD-1 on tumor-infiltrating T cell, delivering inhibitory signal causing exhaustion phenotype (reduced proliferation, cytokine production, killing), with similar pathway shown for CTLA-4/B7 interaction. Panel C: Immunosuppressive cytokine network showing tumor cell secreting TGF-beta (suppressing T cells, promoting Tregs), IL-10 (inhibiting DCs), VEGF (abnormal vasculature, DC suppression), IDO (tryptophan depletion), and adenosine production via CD39/CD73 (suppressing effectors), all converging to create hostile microenvironment. Panel D: Metabolic competition illustration showing tumor cells consuming glucose (Warburg effect producing lactate, lowering pH), creating hypoxia (stabilizing HIF, increasing adenosine), and depleting amino acids (arginine by arginase, tryptophan by IDO), while T cells attempting to function are metabolically starved and inhibited. </image>

IV. Tumor Microenvironment

The tumor microenvironment (TME) encompasses all non-malignant components surrounding and intermingled with cancer cells, including immune cells, fibroblasts, blood vessels, extracellular matrix, and soluble factors. This complex ecosystem profoundly influences tumor behavior, immune responses, and therapeutic outcomes. Tumor cells do not exist in isolation but actively shape their microenvironment to promote survival, growth, and immune evasion. Understanding the TME has become essential for developing effective immunotherapies and predicting which patients will respond. The composition and organization of the TME vary considerably among different cancer types and even among individual tumors of the same type, contributing to the heterogeneity of responses to immunotherapy.

Multiple immunosuppressive cell populations infiltrate tumors and actively suppress anti-tumor immunity. Regulatory T cells (Tregs) accumulate in many tumors, drawn by chemokines and supported by TGF-beta, where they suppress effector T cell responses through multiple mechanisms including IL-10 and TGF-beta secretion, CTLA-4-mediated inhibition, and IL-2 consumption. Myeloid-derived suppressor cells (MDSCs) are immature myeloid cells that expand dramatically in cancer patients and accumulate in tumors, suppressing T cells through arginase (depleting arginine), inducible nitric oxide synthase (producing reactive nitrogen species), and reactive oxygen species production. Tumor-associated macrophages (TAMs) often polarize toward an M2-like phenotype that promotes tumor growth, angiogenesis, and immune suppression rather than the M1-like anti-tumor phenotype. Tumor-associated neutrophils can similarly adopt a pro-tumor N2 phenotype. The balance between immunosuppressive and anti-tumor immune cells within the TME critically determines whether the immune system can control tumor growth.

Cancer-associated fibroblasts (CAFs) and the tumor vasculature contribute to immune exclusion and suppression. CAFs produce extracellular matrix components that create a physical barrier to T cell infiltration and secrete factors that recruit and activate immunosuppressive cells. The tumor vasculature is typically abnormal, with irregular, leaky vessels that create poor perfusion and hypoxia while inefficiently delivering immune cells to the tumor. Endothelial cells in tumor vessels may lack the adhesion molecules needed for T cell extravasation, and VEGF-mediated effects impair T cell trafficking. The net result is often exclusion of effector T cells from the tumor parenchyma, limiting their access to tumor cells regardless of their activation state.

The concept of "hot" versus "cold" tumors has emerged as a critical paradigm for understanding and predicting immunotherapy responses. Hot tumors, also termed inflamed tumors, are characterized by substantial T cell infiltration within the tumor parenchyma, evidence of ongoing immune activity (such as IFN-gamma signatures), and often PD-L1 expression induced by infiltrating lymphocytes. These tumors have the machinery for immune recognition in place but are held in check by checkpoint mechanisms; they tend to respond well to checkpoint inhibitor immunotherapy. Cold tumors lack significant T cell infiltration and may be further classified as immune-excluded (T cells present at the tumor margin but not infiltrating) or immune-desert (no T cell presence at all). Cold tumors generally respond poorly to checkpoint inhibitors alone, as releasing the brakes on T cells provides little benefit when no T cells are present to release. A major goal of current research is developing strategies to convert cold tumors to hot tumors by promoting T cell priming, enhancing trafficking, and overcoming barriers to infiltration.

<image> Panel A: Tumor microenvironment overview showing tumor cells in center surrounded by diverse cellular and structural components: immune cells (T cells, Tregs, TAMs, MDSCs, NK cells), cancer-associated fibroblasts secreting ECM, blood vessels (abnormal tumor vasculature), and gradient of cytokines, metabolites, and oxygen levels across the TME. Panel B: Immunosuppressive cells in TME showing Treg mechanisms (IL-10, TGF-beta, CTLA-4, IL-2 consumption), MDSCs with arginase, iNOS, and ROS production suppressing T cells, M2-like TAMs producing growth factors and immunosuppressive cytokines, and N2 neutrophils with similar pro-tumor phenotype, each cell type with specific markers indicated. Panel C: Physical barriers to T cell infiltration showing cancer-associated fibroblasts producing dense ECM (collagen network) preventing T cell penetration, abnormal tumor vasculature lacking proper adhesion molecules (ICAM, VCAM) for T cell extravasation, and hypoxic regions with T cell exclusion. Panel D: Hot versus cold tumor comparison showing inflamed (hot) tumor with T cells throughout parenchyma, IFN-gamma signature, PD-L1 expression, and good checkpoint inhibitor response; immune-excluded tumor with T cells trapped at margin by stroma; and immune-desert (cold) tumor with no T cell infiltration, poor response to checkpoint blockade, and strategies to convert cold to hot (STING agonists, radiation, oncolytic virus). </image>

V. Checkpoint Inhibitor Immunotherapy

The discovery that blocking inhibitory immune checkpoint molecules could unleash anti-tumor immunity has revolutionized cancer treatment and earned James Allison and Tasuku Honjo the 2018 Nobel Prize in Physiology or Medicine. Ipilimumab, an anti-CTLA-4 monoclonal antibody, became the first checkpoint inhibitor approved for cancer treatment in 2011 after demonstrating improved survival in metastatic melanoma, a disease previously resistant to most treatments. CTLA-4 blockade is thought to work primarily by enhancing T cell activation in lymph nodes during the priming phase, releasing the brake on tumor-specific T cells early in the immune response. Additionally, anti-CTLA-4 antibodies may deplete tumor-infiltrating Tregs through ADCC, reducing local immunosuppression. Ipilimumab produces durable responses in approximately 20% of melanoma patients, with some responses lasting over a decade.

Anti-PD-1 and anti-PD-L1 antibodies have become the most widely used checkpoint inhibitors, with approvals spanning numerous cancer types. Pembrolizumab and nivolumab, both targeting PD-1, are approved for melanoma, non-small cell lung cancer, renal cell carcinoma, head and neck cancer, Hodgkin lymphoma, urothelial carcinoma, microsatellite-unstable cancers regardless of histology, and many other indications. Atezolizumab, durvalumab, and avelumab target PD-L1 with similar therapeutic activity. The mechanism of PD-1/PD-L1 blockade primarily involves reinvigorating exhausted T cells within the tumor microenvironment, restoring their proliferative capacity and effector functions. Response rates vary by tumor type, generally ranging from 15-40% with single-agent therapy, but responses are often durable, with many patients remaining disease-free years after treatment discontinuation. The PD-1/PD-L1 axis appears to be a more broadly relevant checkpoint than CTLA-4, explaining the wider range of approved indications.

Predictive biomarkers help identify patients most likely to benefit from checkpoint inhibitors, though current markers are imperfect. PD-L1 expression on tumor cells, assessed by immunohistochemistry, correlates with response to PD-1/PD-L1 inhibitors in some cancer types and is used to guide treatment decisions, particularly in lung cancer. However, some PD-L1-negative tumors respond while some PD-L1-positive tumors do not, limiting the marker's utility. Tumor mutational burden (TMB) reflects neoantigen load and correlates with checkpoint inhibitor response; the FDA has approved pembrolizumab for TMB-high solid tumors regardless of cancer type. Microsatellite instability-high (MSI-H) or deficient mismatch repair (dMMR) status identifies tumors with defective DNA repair, high TMB, and excellent response rates to checkpoint inhibitors; this was the first tissue-agnostic biomarker approval. The presence of tumor-infiltrating lymphocytes and gene expression signatures of T cell inflammation also predict response. Composite biomarkers integrating multiple features may improve prediction beyond any single marker.

The combination of CTLA-4 and PD-1 blockade enhances efficacy compared to either agent alone, reflecting their complementary mechanisms, but at the cost of increased toxicity. The combination of ipilimumab and nivolumab is approved for melanoma, renal cell carcinoma, hepatocellular carcinoma, and microsatellite-unstable colorectal cancer, among others. Response rates and durability exceed those of monotherapy in most settings. The rationale for combination is that CTLA-4 blockade enhances T cell priming while PD-1 blockade reinvigorates effector T cells, addressing different rate-limiting steps in the anti-tumor immune response. Newer combinations pair checkpoint inhibitors with chemotherapy, radiation, targeted therapies, and other immunotherapeutic approaches to further improve outcomes. The sequencing and selection of combination regimens remains an active area of research, guided by the biological principles underlying each component.

<image> Panel A: CTLA-4 blockade mechanism showing T cell in lymph node receiving antigen presentation from dendritic cell, with CD28 competing against CTLA-4 for B7 binding; without antibody, CTLA-4 dominates and T cell is not fully activated; with anti-CTLA-4 (ipilimumab), CD28 engagement predominates, T cell is activated and proliferates; plus Treg depletion mechanism via ADCC. Panel B: PD-1/PD-L1 blockade mechanism showing exhausted T cell in tumor microenvironment with PD-1 engaged by tumor cell PD-L1, resulting in suppressed function; anti-PD-1 (pembrolizumab, nivolumab) or anti-PD-L1 (atezolizumab) blocks interaction, T cell reinvigorated with restored cytotoxicity, proliferation, and cytokine production; with list of approved indications. Panel C: Predictive biomarkers for checkpoint inhibitors showing PD-L1 IHC staining on tumor cells (with caveats about imperfect correlation), TMB measurement showing mutation count and correlation with neoantigen load, MSI-H/dMMR testing identifying DNA repair defects, and TIL presence with inflamed gene signature, with sensitivity/specificity limitations noted. Panel D: Combination checkpoint therapy diagram showing complementary mechanisms: CTLA-4 blockade enhancing priming phase in lymph node while PD-1 blockade reinvigorating effector phase in tumor, with improved response rates from combination (ipi/nivo) compared to monotherapy, balance scale showing efficacy versus toxicity tradeoff. </image>

VI. Immune-Related Adverse Events

Immune-related adverse events (irAEs) result from checkpoint inhibitor-induced autoimmune inflammation affecting normal tissues and represent the major toxicity of these otherwise well-tolerated treatments. By releasing brakes on T cell activity, checkpoint inhibitors not only enhance anti-tumor immunity but can also unleash autoreactive T cells that attack healthy organs. Approximately 60-80% of patients receiving checkpoint inhibitors experience some grade of irAE, with 10-20% developing severe (grade 3-4) toxicity requiring treatment interruption and aggressive immunosuppression. irAEs can affect virtually any organ system, with skin, gastrointestinal tract, liver, and endocrine glands being most commonly affected. Combination therapy with anti-CTLA-4 and anti-PD-1 increases both the frequency and severity of irAEs compared to monotherapy.

The clinical manifestations of irAEs vary by organ system and require a high index of suspicion for timely diagnosis and management. Cutaneous toxicity is most common, ranging from maculopapular rash to more serious conditions including bullous dermatoses and Stevens-Johnson syndrome. Gastrointestinal toxicity presents as diarrhea and colitis with abdominal pain, sometimes severe enough to cause perforation; colonoscopy may reveal ulceration resembling inflammatory bowel disease. Hepatitis manifests as elevated transaminases and bilirubin, requiring exclusion of other causes including disease progression. Endocrine toxicities include thyroiditis (often causing transient hyperthyroidism followed by hypothyroidism), hypophysitis (inflammation of the pituitary causing multiple hormone deficiencies), and less commonly primary adrenal insufficiency. Pneumonitis presents with dyspnea, cough, and infiltrates on imaging. Rare but serious toxicities include myocarditis (with significant mortality), myositis, encephalitis, and severe cytopenias.

Management of irAEs follows severity-based algorithms that balance toxicity control with preservation of anti-tumor efficacy. Grade 1 (mild) toxicities can often be managed symptomatically while continuing checkpoint inhibitor therapy with close monitoring. Grade 2 (moderate) toxicities generally warrant holding the checkpoint inhibitor and initiating corticosteroids if symptoms do not improve promptly; treatment can often resume after toxicity resolves. Grade 3 (severe) toxicities require checkpoint inhibitor hold and high-dose corticosteroids (typically 1-2 mg/kg prednisone equivalent); steroid-refractory cases may require additional immunosuppression with infliximab, mycophenolate, or other agents. Grade 4 (life-threatening) toxicities mandate permanent checkpoint inhibitor discontinuation and aggressive immunosuppression. The decision to rechallenge after irAE resolution depends on the severity and type of toxicity, the cancer context, and available alternatives; rechallenge is sometimes successful but carries risk of recurrence or new toxicity.

Several special considerations apply to irAE management and patient selection. Endocrine toxicities, once established, may require lifelong hormone replacement (thyroid hormone, hydrocortisone) even if the checkpoint inhibitor can be resumed. Patients with preexisting autoimmune diseases were initially excluded from clinical trials, but real-world evidence suggests that many can receive checkpoint inhibitors with manageable flares, though close monitoring is essential. Interestingly, some studies suggest that development of irAEs correlates with anti-tumor response, raising questions about whether irAE prophylaxis might compromise efficacy. Education of patients and providers about irAE recognition is critical, as early intervention improves outcomes and delays can lead to serious complications including death. The emerging field of onco-rheumatology addresses the management of immune-mediated complications of cancer immunotherapy.

<image> Panel A: irAE mechanism diagram showing checkpoint blockade releasing T cell inhibition, with dual effects: enhanced anti-tumor immunity (beneficial) and unmasked autoreactivity against normal tissues (irAE), with examples of tissue-specific autoantigens in thyroid, colon, lung, and other organs being targeted by reinvigorated T cells. Panel B: Organ-specific irAE manifestations showing skin (rash, vitiligo, bullae), GI tract (colitis with ulceration on colonoscopy image), liver (hepatitis with elevated LFTs), endocrine (thyroiditis, hypophysitis, adrenal insufficiency hormone panel), lung (pneumonitis CT image), heart (myocarditis with concerning symptoms), and CNS (encephalitis, neuropathy), with frequencies indicated. Panel C: irAE management algorithm showing grade-based approach: Grade 1 (continue treatment, symptomatic management, monitor), Grade 2 (hold treatment, consider steroids, resume when resolved), Grade 3 (hold treatment, high-dose steroids, add infliximab or other if refractory), Grade 4 (permanent discontinuation, aggressive immunosuppression), with steroid taper schedule. Panel D: Special considerations showing endocrine irAE requiring permanent hormone replacement (thyroid medication pill, hydrocortisone), preexisting autoimmune disease management (flare risk with monitoring strategy), irAE-response correlation graph suggesting possible efficacy link, and patient education importance with warning signs list. </image>

VII. Adoptive Cell Therapy

Tumor-infiltrating lymphocyte (TIL) therapy represents the earliest form of adoptive cell therapy for cancer and has achieved impressive results in melanoma. The approach involves surgically resecting tumor tissue, isolating the lymphocytes that have naturally infiltrated the tumor, expanding them ex vivo to billions of cells over several weeks, and reinfusing them into the patient following lymphodepleting chemotherapy. The rationale is that TILs have already demonstrated tumor-recognition capacity by trafficking to and accumulating within the tumor, so expanding and reinfusing them at massive numbers can overcome the immunosuppressive barriers that limited their efficacy in vivo. Lymphodepletion with cyclophosphamide and fludarabine before TIL infusion eliminates regulatory cells and creates space and cytokine availability for the infused cells to expand. Lifileucel became the first FDA-approved TIL product in 2024 for advanced melanoma, achieving durable complete responses in a subset of patients who had progressed on prior therapies.

Chimeric antigen receptor (CAR) T cells represent a revolutionary approach that redirects T cells to recognize tumor cells through an engineered receptor rather than the natural T cell receptor. The CAR construct fuses an extracellular antigen-recognition domain (typically a single-chain variable fragment derived from an antibody) to intracellular signaling domains that activate the T cell upon antigen binding. First-generation CARs contained only the CD3-zeta activation domain, providing insufficient signaling for sustained anti-tumor activity. Second-generation CARs added a costimulatory domain (CD28 or 4-1BB), dramatically improving persistence and efficacy. Third-generation CARs incorporate two costimulatory domains. The manufacturing process involves collecting patient T cells by leukapheresis, transducing them with viral vectors encoding the CAR, expanding the engineered cells, and infusing them back into the lymphodepleted patient.

CAR-T therapy has achieved remarkable success against B cell malignancies by targeting CD19, a surface molecule expressed on essentially all B cells including malignant B cells. Multiple CD19-directed CAR-T products are approved for relapsed or refractory B-cell acute lymphoblastic leukemia (B-ALL) and various B-cell lymphomas, including tisagenlecleucel, axicabtagene ciloleucel, lisocabtagene maraleucel, and brexucabtagene autoleucel. Response rates exceed 80% in some settings, with substantial proportions achieving durable complete remissions. B-cell maturation antigen (BCMA)-targeted CAR-T products, including idecabtagene vicleucel and ciltacabtagene autoleucel, are approved for relapsed multiple myeloma with similarly impressive results. The success in hematological malignancies has not yet been replicated in solid tumors, where challenges include identifying appropriate targets, CAR-T trafficking to and penetration of tumors, and the immunosuppressive tumor microenvironment.

The manufacturing and administration of CAR-T therapy involves significant logistical complexity and specialized expertise. Production requires three to four weeks from leukapheresis to product release, during which patients may require bridging therapy to control their disease. The lymphodepleting chemotherapy regimen causes cytopenias and infection risk. CAR-T products require careful handling and must be infused at certified treatment centers with capability to manage the unique toxicities of this therapy. The cost of CAR-T products exceeds $400,000 for the cells alone, not including hospitalization, toxicity management, and supportive care. Despite these challenges, CAR-T therapy has become standard of care for certain relapsed hematological malignancies and is extending to earlier lines of therapy as data mature.

<image> Panel A: TIL therapy process showing surgical tumor resection, TIL isolation and ex vivo expansion over several weeks (billions of cells in culture flasks), lymphodepleting chemotherapy (cyclophosphamide/fludarabine), TIL infusion, and durable response in melanoma patient, with rationale that tumor-reactive T cells are selected by natural infiltration and massively expanded. Panel B: CAR structure evolution showing first-generation CAR (scFv extracellular domain + CD3-zeta only, insufficient persistence), second-generation (scFv + CD28 or 4-1BB + CD3-zeta, improved efficacy), and third-generation (two costimulatory domains), with detailed diagram of each component and its function in T cell activation. Panel C: CAR-T manufacturing workflow showing patient leukapheresis collection, T cell isolation, viral transduction with CAR vector, ex vivo expansion (3-4 weeks), quality control testing, lymphodepleting chemotherapy, and CAR-T infusion, with timeline and logistics considerations including bridging therapy during manufacturing. Panel D: Approved CAR-T products table showing CD19-targeted products (tisagenlecleucel, axicabtagene ciloleucel, lisocabtagene, brexucabtagene) for B-ALL and lymphomas with response rates, and BCMA-targeted products (idecabtagene, ciltacabtagene) for multiple myeloma, plus notation of solid tumor challenges. </image>

VIII. CAR-T Toxicities

Cytokine release syndrome (CRS) is the most common and characteristic toxicity of CAR-T therapy, resulting from massive activation of the infused CAR-T cells and subsequent release of inflammatory cytokines. When CAR-T cells encounter and engage their target antigen, they proliferate rapidly and release large quantities of cytokines including IL-6, IFN-gamma, and IL-1, which activate macrophages and other immune cells, amplifying the cytokine cascade. Clinical manifestations typically begin within the first week after infusion and range from mild fever and fatigue to severe multiorgan dysfunction resembling septic shock. Fever is universal and often the first sign; progression may include hypotension requiring vasopressors, hypoxia requiring supplemental oxygen or mechanical ventilation, and end-organ dysfunction affecting kidneys, liver, and heart. CRS severity is graded from 1 (fever alone) to 4 (life-threatening organ dysfunction requiring ICU-level support).

Management of CRS employs tocilizumab, an IL-6 receptor antagonist, as the cornerstone of treatment. IL-6 is a key driver of CRS pathophysiology, and blocking its receptor rapidly ameliorates symptoms in most cases. Tocilizumab is typically administered when CRS reaches grade 2 or higher, though practice varies. Corticosteroids are reserved for tocilizumab-refractory CRS or severe toxicity, as there is theoretical concern that steroids might compromise CAR-T efficacy by suppressing the anti-tumor immune response; however, this concern has not been consistently borne out in clinical studies. Supportive care including fluids, vasopressors, and oxygen are essential. Most CRS resolves within one to two weeks with appropriate management, though severe cases can be prolonged. Prophylactic tocilizumab is being studied to prevent or mitigate CRS without compromising efficacy.

Immune effector cell-associated neurotoxicity syndrome (ICANS) is a unique neurological complication of CAR-T therapy that often follows or coincides with CRS. The pathophysiology is incompletely understood but appears to involve cytokine-mediated disruption of the blood-brain barrier, allowing inflammatory mediators and potentially CAR-T cells to enter the central nervous system. Clinical features include word-finding difficulty and aphasia (often the earliest signs), confusion and disorientation, tremor, headache, and in severe cases, seizures, cerebral edema, and obtundation. ICANS is graded using the Immune Effector Cell-Associated Encephalopathy (ICE) score, which assesses orientation, naming, following commands, writing, and attention. Unlike CRS, ICANS responds poorly to tocilizumab (likely because IL-6 receptor blockade does not address CNS pathology); corticosteroids are the primary treatment. Most cases resolve completely, but severe ICANS can cause lasting neurological deficits or death.

Additional complications of CAR-T therapy require vigilance and proactive management. B-cell aplasia is an expected on-target, off-tumor effect of CD19-directed CAR-T therapy, as normal B cells also express CD19; this results in hypogammaglobulinemia requiring immunoglobulin replacement in many patients. Prolonged cytopenias, affecting neutrophils, platelets, or all lineages, can persist for months and increase infection risk. Opportunistic infections, particularly viral reactivation (CMV, HSV, EBV) and Pneumocystis jirovecii, require prophylaxis and monitoring. Rare cases of secondary malignancies, including myelodysplastic syndrome and T-cell malignancies containing the CAR construct, have been reported and are under investigation. The requirement for specialized toxicity management has led to REMS (Risk Evaluation and Mitigation Strategy) programs mandating that CAR-T products be administered only at certified centers with appropriate expertise and resources.

<image> Panel A: CRS pathophysiology showing CAR-T cell engaging target antigen on tumor cell, rapid proliferation and activation, release of IFN-gamma and TNF-alpha, macrophage activation with IL-6, IL-1 production, cytokine cascade amplification, and systemic effects (fever, hypotension, hypoxia, organ dysfunction) with severity grading scale. Panel B: CRS management algorithm showing supportive care foundation (fluids, vasopressors, oxygen), tocilizumab as first-line specific treatment blocking IL-6 receptor, corticosteroids for refractory cases, with decision tree based on CRS grade and response to initial interventions, and typical resolution timeline. Panel C: ICANS pathophysiology and presentation showing blood-brain barrier disruption from systemic inflammation, neuroinflammation with cytokines and possibly CAR-T cells in CNS, clinical features (word-finding difficulty, confusion, tremor, seizures, cerebral edema), ICE scoring assessment, and corticosteroid-based treatment approach. Panel D: Other CAR-T complications showing B-cell aplasia from CD19 targeting (both tumor and normal B cells depleted, hypogammaglobulinemia requiring IVIG), prolonged cytopenias with infection risk, opportunistic infection prophylaxis recommendations, and secondary malignancy concerns with ongoing monitoring. </image>

IX. Other Immunotherapies

Cancer vaccines aim to induce or enhance tumor-specific immune responses by presenting tumor antigens to the immune system in an immunogenic context. Therapeutic cancer vaccines differ from preventive vaccines (such as HPV vaccines) in that they must overcome tolerance and immunosuppression in patients with established cancer. Sipuleucel-T, the first FDA-approved therapeutic cancer vaccine, is an autologous cellular immunotherapy for metastatic prostate cancer; patient's dendritic cells are exposed ex vivo to a fusion protein of prostatic acid phosphatase (a prostate-specific antigen) and GM-CSF, then reinfused. Personalized neoantigen vaccines represent a cutting-edge approach that identifies patient-specific mutational neoantigens through tumor sequencing, predicts which are most likely to be immunogenic, synthesizes corresponding peptides or encodes them in mRNA, and administers them with adjuvant to prime neoantigen-specific T cells. Early clinical trials of personalized vaccines combined with checkpoint inhibitors have shown promising results in melanoma and other tumors.

Oncolytic viruses are naturally occurring or engineered viruses that selectively infect and replicate within tumor cells, causing oncolysis and releasing tumor antigens to stimulate anti-tumor immunity. Talimogene laherparepvec (T-VEC), a modified herpes simplex virus expressing GM-CSF, is approved for injectable melanoma and represents the first oncolytic virus approved in Western markets. The virus preferentially replicates in tumor cells lacking antiviral type I interferon responses, leading to lysis of infected cells. Beyond direct tumor killing, the released tumor antigens, danger signals, and GM-CSF promote dendritic cell activation and tumor-specific T cell responses, potentially creating abscopal effects on uninjected tumors. Combining oncolytic viruses with checkpoint inhibitors may convert immunologically cold tumors to hot tumors by promoting inflammation and T cell infiltration. Multiple oncolytic virus platforms are in clinical development.

Bispecific antibodies engage two different targets simultaneously, enabling novel therapeutic mechanisms. Bispecific T cell engagers (BiTEs) bridge T cells to tumor cells by simultaneously binding CD3 on T cells and a tumor antigen, forcing T cell activation and tumor cell killing regardless of T cell receptor specificity or MHC expression. Blinatumomab, a CD19xCD3 BiTE, is approved for B-ALL and represents proof of concept for this approach. Like CAR-T cells, BiTEs can cause CRS and neurotoxicity, requiring specialized management. Other bispecific formats include T cell-engaging antibodies targeting various tumor antigens and bispecifics that simultaneously block two checkpoint molecules or deliver payloads to tumors. The bispecific antibody field is rapidly expanding, with numerous agents in development across cancer types.

Cytokine therapies were among the earliest immunotherapies for cancer but have largely been supplanted by checkpoint inhibitors and cellular therapies due to toxicity and limited efficacy. High-dose interleukin-2 (IL-2) achieved durable complete responses in a small subset of patients with metastatic melanoma and renal cell carcinoma but causes severe capillary leak syndrome and requires ICU-level monitoring. Interferon-alpha was used as adjuvant therapy for high-risk melanoma and for treatment of hairy cell leukemia and chronic myelogenous leukemia but has significant side effects including flu-like symptoms, depression, and autoimmune phenomena. Current research focuses on engineering cytokines with improved selectivity for anti-tumor immune cells or tumor-localized activity to achieve therapeutic benefit without systemic toxicity. Fusion proteins linking cytokines to tumor-targeting antibodies and next-generation IL-2 variants that preferentially activate effector T cells over regulatory T cells are in clinical development.

<image> Panel A: Cancer vaccine approaches showing sipuleucel-T process (patient DCs exposed to PAP-GM-CSF fusion ex vivo, reinfused), personalized neoantigen vaccine pipeline (tumor sequencing, neoantigen prediction, peptide/mRNA synthesis, patient vaccination with checkpoint inhibitor combination), and DC vaccine concept with tumor antigen loading. Panel B: Oncolytic virus mechanism showing T-VEC (modified HSV) selectively infecting tumor cell (exploiting defective interferon response), viral replication and oncolysis, release of tumor antigens and DAMPs, GM-CSF expression recruiting DCs, priming of tumor-specific T cells, and potential abscopal effect on distant tumors. Panel C: Bispecific antibody structure and function showing BiTE (blinatumomab) with anti-CD3 arm engaging T cell and anti-CD19 arm engaging B cell tumor, bridging cells to force immunological synapse and T cell-mediated killing regardless of TCR specificity, with comparison to CAR-T approach and toxicity profile (CRS, neurotoxicity). Panel D: Cytokine therapy evolution showing historical IL-2 mechanism (T cell proliferation and activation with severe capillary leak toxicity), IFN-alpha uses and limitations, and next-generation approaches (tumor-targeted IL-2 fusion proteins, IL-2 variants preferentially activating CD8 effectors over Tregs) aiming for efficacy without systemic toxicity. </image>

X. Combination Strategies and Future Directions

Combination immunotherapy strategies aim to address the multiple barriers to effective anti-tumor immunity by targeting complementary pathways. The combination of anti-CTLA-4 and anti-PD-1 antibodies has already demonstrated superior efficacy compared to either agent alone in multiple tumor types, reflecting their distinct mechanisms of action in T cell priming versus effector function. Combining checkpoint inhibitors with chemotherapy has become standard of care in lung cancer and triple-negative breast cancer; chemotherapy may enhance immunogenicity by causing immunogenic cell death that releases tumor antigens and danger signals. Checkpoint inhibitors combined with targeted therapies (such as BRAF/MEK inhibitors in melanoma or VEGF inhibitors in renal cell carcinoma) can address both cell-intrinsic oncogenic signaling and immune evasion. Radiation therapy causes immunogenic cell death and may synergize with checkpoint blockade to induce abscopal responses in unirradiated tumors.

Overcoming resistance to immunotherapy represents a major focus of current research. Primary resistance, where tumors never respond to immunotherapy, and acquired resistance, where initially responding tumors progress, both limit the benefit of current treatments. Strategies to convert cold tumors to hot tumors include intratumoral injection of immune-stimulating agents (STING agonists, TLR agonists, oncolytic viruses), radiation therapy to induce immunogenic cell death, and interventions to enhance T cell trafficking and infiltration. Targeting additional checkpoint molecules beyond PD-1 and CTLA-4 addresses T cell exhaustion that may persist despite dual checkpoint blockade; inhibitors of LAG-3, TIM-3, and TIGIT are in clinical development, with LAG-3 inhibitor relatlimab now approved in combination with nivolumab for melanoma. Modulating the tumor microenvironment by depleting MDSCs, reprogramming macrophages, or targeting metabolic barriers may also enhance immunotherapy efficacy.

Novel targets and modalities continue to expand the immunotherapy landscape. CD47, the "don't eat me" signal that protects cells from phagocytosis, is expressed at high levels on many tumor cells; blocking CD47-SIRPalpha interaction enables macrophage-mediated phagocytosis of tumor cells. STING (stimulator of interferon genes) agonists activate innate immunity and can convert cold tumors to hot tumors by inducing type I interferon production and promoting T cell priming. Adenosine pathway inhibitors block the immunosuppressive effects of adenosine in the tumor microenvironment. Allogeneic "off-the-shelf" CAR-T products, derived from healthy donor T cells with modifications to prevent graft-versus-host disease and rejection, could overcome the logistical challenges and delays of autologous CAR-T manufacturing. CAR-NK cells, which may cause less cytokine release syndrome and do not require HLA matching, are in clinical development.

The future of cancer immunotherapy points toward increasingly personalized and combination approaches guided by biomarkers and real-time monitoring. Comprehensive molecular profiling of tumors will identify the most appropriate immunotherapy strategies for individual patients based on tumor antigen landscape, immune contexture, and resistance mechanisms. Liquid biopsies monitoring circulating tumor DNA may enable early detection of response or resistance, guiding treatment adaptations. Next-generation cellular therapies with armored CARs (secreting cytokines or checkpoint-blocking molecules), logic-gated CARs (requiring multiple antigens for activation), and enhanced persistence may overcome current limitations. Gene editing technologies including CRISPR are enabling more precise engineering of cellular therapies. The integration of immunotherapy with other treatment modalities, informed by deeper understanding of tumor-immune interactions, promises continued improvement in outcomes for patients with cancer.

<image> Panel A: Combination immunotherapy rationale showing multiple barriers to anti-tumor immunity (poor priming, checkpoint inhibition, immunosuppressive TME, physical barriers) and strategies addressing each: checkpoint combo (CTLA-4 + PD-1), chemotherapy immunogenic cell death, radiation abscopal effect, targeted therapy addressing oncogenic signaling, with synergy diagram. Panel B: Overcoming resistance strategies showing cold tumor conversion methods (intratumoral STING agonist injection inducing IFN production and T cell recruitment, oncolytic virus creating inflammation, radiation inducing immunogenic cell death), next-generation checkpoint targets (LAG-3, TIM-3, TIGIT) beyond PD-1/CTLA-4, and TME modulation approaches. Panel C: Novel targets and modalities showing CD47/SIRPalpha blockade enabling macrophage phagocytosis of tumor cells, STING agonist innate immune activation pathway, adenosine pathway inhibition diagram, and allogeneic CAR-T concept (universal donor cells, gene editing to remove TCR and add CAR, off-the-shelf product available immediately). Panel D: Future vision showing comprehensive tumor profiling (genomic, transcriptomic, immunologic) informing personalized immunotherapy selection, liquid biopsy monitoring ctDNA for response assessment, next-generation CAR designs (armored CAR secreting IL-12, logic-gated CAR with dual antigen requirement), and integrated multimodal therapy timeline. </image>

Summary

• Cancer immunoediting describes the three phases of tumor-immune interaction: elimination (immunosurveillance), equilibrium (containment), and escape (clinical cancer)
• Tumor antigens include neoantigens (mutation-derived, highly immunogenic), tumor-associated antigens (overexpressed), viral antigens (in virally-associated cancers), and differentiation antigens
• Tumor immune evasion involves antigen loss, MHC downregulation, checkpoint molecule upregulation (PD-L1), and secretion of immunosuppressive factors (TGF-beta, IL-10, VEGF)
• The tumor microenvironment contains immunosuppressive cells (Tregs, MDSCs, M2 macrophages) and physical barriers; hot tumors (T cell-infiltrated) respond better to immunotherapy than cold tumors
• Checkpoint inhibitors (anti-CTLA-4, anti-PD-1/PD-L1) release brakes on T cell activity; predictive biomarkers include PD-L1, TMB, and MSI-H status
• Immune-related adverse events affect multiple organ systems and require severity-based management with corticosteroids and other immunosuppression
• CAR-T cells are engineered with chimeric antigen receptors targeting CD19 (B cell malignancies) or BCMA (myeloma), achieving high response rates with CRS and ICANS as key toxicities
• TIL therapy expands naturally tumor-infiltrating lymphocytes for reinfusion, with FDA approval for melanoma (lifileucel)
• Other immunotherapies include cancer vaccines, oncolytic viruses, bispecific antibodies, and cytokine therapies
• Future directions include combination strategies, novel checkpoint targets, approaches to convert cold tumors to hot, and next-generation cellular therapies

Key Terms

This content is subject to the MIT License. © 2024–2026 Hibbert School of Medicine.

Read this lecture as Markdown