Mednotes

L5 Cancer Principles of non-chemotherapy systemic anti-cancer treatments

Properties2
tagscancer, lecture
aliasesL5 Cancer Principles of non-chemotherapy systemic anti-cancer treatments

26 min read

Overview

This lecture covers systemic anti-cancer therapy other than cytotoxic chemotherapy: hormonal therapy, small-molecule targeted agents, monoclonal antibodies and immunotherapy. It sets the scene with the burden of cancer in New Zealand and the tumour heterogeneity that makes “one cancer really many cancers”, grounds drug action in the cell cycle and in the generic ligand/receptor/kinase architecture of signalling pathways, then works through each drug class with its nomenclature, toxicities and evidence. Two extended case studies carry the argument: EGFR inhibition in colorectal and gastric cancer, where predictive biomarkers and trial context decide whether a drug helps or harms, and melanoma, where BRAF inhibition and checkpoint blockade are set against resistance and brain metastases.

Burden of cancer in New Zealand

  • Cancer is New Zealand’s leading cause of death, at 29% of deaths. The remainder of the causes-of-death pie: other 26%, ischaemic heart disease 18%, cerebrovascular disease 9%, respiratory disease 6%, other heart disease 4%, diabetes 3%, pneumonia/influenza 2%, suicide 2%, motor vehicle accidents 1%, assault 0%.
  • Five types account for the majority of cases: prostate 14%, breast 14%, colorectal 14%, melanoma 11%, lung 9%. Then other 25%, non-Hodgkins 4%, leukaemia 3%, uterus 2%, kidney 2%, pancreas 2%.
  • The number of cancer deaths is rising: from under 1,000 per year in 1948 to roughly 9,000 per year by about 2012.
  • The death rate is falling: deaths per 100,000 population rose from about 135 in 1948 to a peak of about 165 around 1988 to 1990, then declined to about 125 by 2012.
  • The core problem stated: things that look the same do not behave the same, so cancers of the same name differ. One cancer is really many cancers, the tumour heterogeneity and resistance framing of Dagogo-Jack and Shaw (Nat Rev Clin Oncol, 2017).

Cellular replication as the basis of treatment

  • Cancer is a disease of uncontrolled replication.
  • Cell cycle: interphase, then G1 (cellular contents other than the chromosomes are duplicated), S (each of the 46 chromosomes is duplicated), G2 (the cell double checks the duplicated chromosomes for error and makes any needed repairs), mitosis, cytokinesis, with G0 as cell cycle arrest. Transition between stages is regulated by cell dependent kinases (CDKs), as the slide names them.
  • Mitosis in order, with the defining events given:
    • Prophase: nuclear membrane breaks down, loosely coiled replicated chromosomes condense, sister chromatids joined at the centromere, spindle fibres appear.
    • Prometaphase: spindle fibres attach to chromosomes.
    • Metaphase: chromosomes condense and align.
    • Anaphase: centromeres divide and sister chromatids move to opposite poles.
    • Telophase: nuclear membrane reforms, chromosomes decondense, spindle fibres disappear.
    • Cytokinesis: cytoplasm divides and the parent cell becomes two daughter cells with identical genetic information.

Classes of systemic anti-cancer therapy and where they act

Medical oncology systemic anti-cancer therapy comprises five categories: chemotherapy, hormonal therapies, targeted therapies, monoclonal antibodies and immunotherapy.

Mechanisms of cancer therapies, as listed:

  • Target cellular replication, for example chemotherapy and hormones.
  • Target protein synthesis and function, for example protein kinase inhibitors (tyrosine kinase inhibitors).
  • Target extracellular signalling protein, for example monoclonal antibodies such as EGFR and HER2. The slide highlights this bullet as the definition of “targeted therapies”.
  • Target ligand for extracellular signalling proteins.
  • Antibodies against cell-surface components.
  • Immune therapies.

Architecture of targeted therapy

  • Every signalling pathway has a ligand, a transmembrane protein and an intracellular tyrosine kinase before it “talks” to the nucleus.
  • There are monoclonal antibodies against the ligand and against the extracellular domain of the protein. There are tyrosine kinase inhibitors for the intracellular proteins.
  • The main pathways are VEGF, EGF and HER2 (EGFR2).
  • The recurring pathway schematic runs: ligand to EGFR to Ras to RAF to MEK, with the output arrow to cell growth, proliferation, invasion and metastases. Successive builds show blockade of the ligand, blockade of EGFR, an antibody bound to the extracellular EGFR domain, and a prohibition symbol placed on RAF, so the three illustrated intervention points are the ligand, the receptor and the intracellular kinase node.
  • A pathway figure from Wadhwa et al. (gastric cancer molecular and clinical dimensions, Nat Rev Clin Oncol 2013) maps agents onto their targets: bevacizumab on VEGF (40%); ramucirumab on VEGFR (36%); cetuximab, matuzumab, panitumumab and nimotuzumab on EGFR/HER1 (27-44%); trastuzumab and pertuzumab on HER2 (7-34%); rilotumumab and onartuzumab on HGF, and crizotinib on c-MET (10-15%); figitumumab on IGFR; dovitinib, sunitinib and sorafenib on FGFR2 (9%), PDGFR and VEGFR; erlotinib and lapatinib at the EGFR/HER2 kinase; intracellular Ras (2-20%), Raf1 (0-3%, blocked by sorafenib and regorafenib), Mek, Erk; PI3K (4-36%), PDK-1, Akt (29-86%, GCD-0068, MK 0226, perifosine), mTOR (60-88%, everolimus); Hh (16%)/Ptch-1 with LDE 225, BMS833923 and GDC-0449; bortezomib on the proteasome; marimastat; vorinostat on HDAC; chromatin-remodelling gene (EZH2, ARID, KMT2C, KMT2A) inhibitors; outputs proliferation, survival/apoptosis, metastasis and angiogenesis.

Warning

The crizotinib pathway figure is small and dense in the source: the right-hand edge of the figure is partly covered by a light-bulb icon.

Rapid development of tyrosine kinase inhibitors

A 2001 to 2020 approval timeline shows how fast the class grew, colour coded by target class: EGFR-TKI, ALK-TKI, TRK-TKI, HER2-TKI, VEGFR-associated multi-targeted TKI, RET-TKI, MET-TKI, MEK-TKI, FGFR-TKI and KIT/PDGFR-TKI.

By year: 2001 imatinib; 2002 gefitinib; 2004 erlotinib; 2005 sorafenib; 2006 sunitinib; 2007 lapatinib; 2009 pazopanib; 2011 icotinib, crizotinib, vandetanib; 2012 axitinib, cabozantinib, regorafenib; 2013 trametinib, afatinib; 2014 apatinib, alectinib, ceritinib; 2015 cobimetinib, lenvatinib, osimertinib; 2017 neratinib, brigatinib; 2018 pyrotinib, binimetinib, anlotinib, fruquintinib, lorlatinib, larotrectinib, dacomitinib; 2019 erdafitinib, entrectinib; 2020 tepotinib, capmatinib, pemigatinib, ripretinib, avapritinib, tucatinib, selumetinib, pralsetinib, selpercatinib, almonertinib.

Footnote key: one asterisk is a first-generation TKI, two asterisks second-generation, three asterisks third-generation; ”&” marks first approval by the NMPA and the section mark first approval by the MHLW, with all other drugs first approved by the FDA.

Hormonal therapy: the original targeted therapy

  • The lecture calls hormonal therapy the original targeted therapy, and reproduces the title page of a historical paper: “On the treatment of inoperable cases of carcinoma of the mamma: suggestions for a new method of treatment, with illustrative cases”, by George Thomas Beatson, M.D. Edin., published in The Lancet on 11 July 1896. [slide shows the title page only and does not state what the new method was]
  • The oestrogen receptor was identified in breast cancer in 1968.
  • 1973: laboratory experiments identify Tamoxifen as an antagonist in ER positive breast cancer.
  • 1982: adjuvant tamoxifen in ER positive breast cancer is shown to reduce relapse.
  • The point drawn from this sequence is that progress is not always linear.

GIST and the start of the modern targeted era

The lecture marks the “birth of the targeted therapy era/error” with the brief report “Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor”, N Engl J Med, volume 344, number 14, 5 April 2001. Paired whole-body scans show multiple tumour foci before treatment and resolution of most foci afterwards.

Monoclonal antibodies

Nomenclature follows the degree of humanisation, and the suffix tells you where an antibody sits on that spectrum:

  • Mouse: 100% mouse protein.
  • Chimeric: 34% mouse protein, suffix “-uximab”.
  • Humanised: 10% mouse protein, suffix “-uzumab”.
  • Fully human: 100% human protein, suffix “-umumab”.

The two associated effects flagged alongside the spectrum are infusion reactions and ADCC (antibody-dependent cellular cytotoxicity).

DrugStructureTargetIndications
RituximabChimericCD-20 receptorB-cell lymphoma
CetuximabChimericEGFR receptorHead and neck, colorectal
TrastuzumabHumanisedHER-2 receptorBreast, gastric
PanitumumabHumanEGFR receptorColorectal
BevacizumabHumanisedVEGF ligandColorectal, ovary, CNS
RamucirumabHumanVEGF receptorGastric

Drug-antibody conjugates

T-DM1 is the worked example, presented as a delivery vehicle for a cytotoxic:

  • Composition: DM1 (3 to 4 molecules per IgG), a thioether linker, and trastuzumab (HzIgG1, attached at LysNH2 residues at random).
  • Antibody-drug conjugate mechanism in five steps: (1) ADC circulates in plasma, (2) ADC binds to its receptor, (3) the ADC-receptor complex is internalised, (4) the cytotoxic agent is released, (5) apoptosis of the target cell.

Toxicities of targeted therapies

Toxicity may be on target or off target. Categories given:

  • Cardiovascular: hypertension, LVEF, VTE, QTc.
  • Gastrointestinal: nausea, diarrhoea, appetite.
  • Skin.
  • Renal: glomerular, tubular, interstitial.

Anti-VEGF TKI hypertension, an on-target effect

Three mechanisms:

  1. Impaired balance between vasoconstrictors and vasodilators: decreased nitric oxide production in arteriolar walls, decreased prostaglandin I2 (prostacyclin), increased endothelin-1.
  2. Capillary rarefaction: decreased endothelial cell survival from VEGF inhibition, raising systemic vascular resistance.
  3. Renal effects of VEGF inhibition: VEGF is expressed by podocytes and VEGFR by glomerular mesangial and endothelial cells, so inhibition reduces mesangial and endothelial cell survival, disrupts the glomerular filtration barrier, damages the renal vasculature and decreases glomerular filtration rate.

Mean blood pressure rise from baseline on treatment, by drug:

DrugIncrease in SBP (mmHg)Increase in DBP (mmHg)
Sunitinib10.17.6
Sorafenib8.95.9
Pazopanib7.56.9
Cabozantinib4.26.7
Axitinib12.610.3

Skin toxicity

Paired before and after clinical photographs show an erythematous papulopustular rash over the face, neck and shoulders as an example of drug-induced skin toxicity.

Case study: EGFR inhibition

CRYSTAL: cetuximab added to FOLFIRI

  • Primary endpoint progression-free survival, ITT population, independent review, 599 patients per arm.
  • PFS ITT hazard ratio 0.85, p = 0.048. Median PFS 8.9 months with cetuximab plus FOLFIRI versus 8.0 months with FOLFIRI. The slide records a 1-year PFS rate of 23% versus 34%.

KRAS status as the predictive biomarker

  • In the cetuximab plus FOLFIRI arm the effect is confined to KRAS wild-type disease: hazard ratio 0.63, p = 0.007, median PFS 9.9 months in wild-type (n = 172) versus 7.6 months in mutant (n = 105).
  • In the FOLFIRI-alone arm KRAS status makes no difference: hazard ratio 0.97, p = 0.87, median PFS 8.7 months wild-type (n = 176) versus 8.1 months mutant (n = 87), and the two curves do not separate.

Important

KRAS status separates outcomes only in the arm that contains the antibody.

New EPOC: the same drug can harm in a different setting

  • Design: patients with operable, including borderline operable, colorectal liver metastases were randomised. Arm A (control) received chemotherapy for 12 weeks, liver resection, then chemotherapy for 12 weeks. Arm B (experimental) received chemotherapy plus cetuximab for 12 weeks, liver resection, then chemotherapy plus cetuximab for 12 weeks. Presented by John Neil Primrose at the 2013 ASCO Annual Meeting.
  • Primary analysis in all randomised KRAS wild-type patients: hazard ratio 1.49, 95% CI 1.04 to 2.12, p = 0.030. Median PFS 20.5 months in Arm A versus 14.1 months in Arm B, so adding cetuximab in the potentially curable setting was worse despite KRAS wild-type selection.

Gastric cancer and REAL3

  • Rationale: EGFR is over-expressed in 27 to 55% of gastric cancers, over-expression is associated with worse prognosis, and rates of K-ras mutation are low. Standard of care chemotherapy included oxaliplatin and a fluoropyrimidine, so the trial question was EOC with or without panitumumab.
  • REAL3 (Waddell, Chau, Cunningham and colleagues; Lancet Oncology 2013;14(6):481-489) randomised previously untreated advanced oesophagogastric cancer to epirubicin, oxaliplatin and capecitabine (EOC) or modified EOC plus panitumumab (mEOC+P). Overall survival in the 553-patient intention-to-treat population: hazard ratio 1.37, 95% CI 1.07 to 1.76, p = 0.013, again against the antibody arm. Patients still on treatment at trial closure and those crossing over were censored.

Lessons from EGFR

  • Over-expression or amplification of a pathway is a hint, not a home run.
  • Treatment effect may be specific to tumour stage.
  • Interaction with other drugs is also important.
  • Site of primary origin may impact on efficacy.
  • Most tumours likely have multiple drivers rather than a single driver.

Immunotherapy: the checkpoint landscape

Immune checkpoint signalling is both co-inhibitory and co-stimulatory, with red arrows for inhibitory and green for stimulatory effects on the reproduced figure.

  • Tumour cell to T cell or NK cell pairs: HLA-E with CD94/NKG2A, HLA-C with KIR, CD155/CD112/CD113 with TIGIT, CD155/CD111 with CD96, MHC-II with LAG3, CEACAM-1 and galectin 9 with TIM-3, IDO/TDO and adenosine as inhibitory signals, and PD-L1 with PD-1.
  • Antigen-presenting cell to T cell pairs: TIM-3 with galectin 9, GITR with GITRL, CD40 with CD40L, OX40 with OX40L, 4-1BB with 4-1BBL, TIGIT with CD155, ICOS with ICOSL, LAG3 with MHC-II, TCR with MHC, CD28 and CTLA4 with CD80/CD86, PD-1 with PD-L1/PD-L2, CD27 with CD70, plus VISTA and CD47 with SIRP-alpha.

Warning

The checkpoint figure is dense and some of the small paired labels sit at the edge of legibility at the rendered resolution.

Anti-CTLA-4

The diagram shows a dendritic cell presenting to a T cell through MHC with TCR and CD86 with CD28, while CTLA-4 competes at that contact. An anti-CTLA-4 antibody blocks the CTLA-4 interaction, and the right of the slide shows the activated T cell attacking a tumour mass. The contrast drawn is between T cell activation leading to tumour killing and inhibition of activation leading to no killing.

Anti-PD-1 and anti-PD-L1

The diagram shows a tumour cell and T cell engaged through MHC with TCR while PD-L1 binds PD-1. Anti-PD-L1 and anti-PD-1 antibodies block that interaction, restoring the T cell attack on the tumour. The contrast drawn is between inhibition of killing giving no response and relief of that inhibition giving tumour killing.

Warning

On both the anti-CTLA-4 and anti-PD-1/PD-L1 slides the title-area text is two overlapping sentences and is rendered garbled, so the wording above is approximate.

Immunotherapy toxicities

Immune-related adverse events span essentially every organ system:

  • Eye: uveitis, conjunctivitis, scleritis or episcleritis, blepharitis, retinitis.
  • Endocrine: hyperthyroidism or hypothyroidism, hypophysitis, adrenal insufficiency, diabetes.
  • Respiratory: pneumonitis, pleuritis, sarcoid-like granulomatosis.
  • Liver: hepatitis.
  • Cardiovascular: myocarditis, pericarditis, vasculitis.
  • Renal: nephritis.
  • Gastrointestinal: colitis, ileitis, pancreatitis, gastritis.
  • Skin: rash, pruritus, psoriasis, vitiligo, DRESS, Stevens Johnson.
  • Neurologic: neuropathy, Guillain Barre, myelopathy, meningitis, encephalitis, myasthenia.
  • Blood: haemolytic anaemia, thrombocytopenia, neutropenia, haemophilia.
  • Musculoskeletal: arthritis, dermatomyositis.

Resistance to immunotherapy

Four clinical scenarios (Sharma et al, Cell 2017):

  • A: the tumour is resistant to immunotherapy with no active immune response.
  • B: the tumour is resistant despite an active anti-tumour immune response, which is turned off by checkpoints or other adaptive resistance mechanisms.
  • C: initial response then progression, caused by a heterogeneous tumour population and selection of pre-existing resistant clones.
  • D: initial response then progression through true acquired resistance developing during immunotherapy.

Mechanisms of primary and adaptive resistance:

  • Tumour cell intrinsic: absence of antigenic proteins (low mutational burden, lack of viral antigens, lack of cancer-testis antigens, overlapping surface proteins); absence of antigen presentation (deletion in TAP, deletion in B2M, silenced HLA); genetic T cell exclusion (MAPK oncogenic signalling, stabilised beta-catenin, mesenchymal transcriptome, oncogenic PD-L1 expression); insensibility to T cells (mutations in interferon gamma pathway signalling).
  • Tumour cell extrinsic: absence of T cells (lack of T cells with tumour antigen-specific TCRs); inhibitory immune checkpoints (VISTA, LAG-3, TIM-3); immunosuppressive cells (tumour-associated macrophages, Tregs).

Case study: melanoma

Multiple phenotypes, multiple diseases

  • Clinical photographs show the range of presentations: a subungual pigmented lesion of the thumb, a pigmented nodular skin lesion, a large irregular pigmented skin lesion, and a pigmented lesion of the oral mucosa or gum. [slides give no individual captions for these images]
  • What looks like one entity, a single circle labelled melanoma, resolves into molecular subtypes: an outer division into BRAF wild-type versus BRAF mutant, and an inner division into CD117 negative versus positive.

Warning

The subtype circle labels overlap on the slide and are partly illegible, with “Braf mutant CD117” also printed centrally.

BRAF inhibition: vemurafenib

  • Vemurafenib in V600E mutated melanoma: response is rapid, within days, and the drug is well tolerated.
  • Toxicities listed: photosensitivity, diarrhoea, hand-foot syndrome (HFS), keratoacanthoma (KA) and squamous cell carcinoma (SCC).
  • A waterfall plot of percent change from baseline in target lesion diameters shows shrinkage across most patients spanning unresectable stage IIIc, M1a, M1b and M1c disease (adapted from Chapman et al, 2011).
  • Duration of tumour control versus dacarbazine: hazard ratio 0.38, 95% CI 0.32 to 0.46, log-rank p < 0.001 (post hoc), with vemurafenib (n = 337) improving median progression-free survival by 5.3 months over dacarbazine (n = 338), 1.6 versus 5.9 months.

Checkpoint blockade: pembrolizumab

Pembrolizumab given every 2 weeks or every 3 weeks is compared with ipilimumab for progression-free survival and overall survival, with per-patient waterfall plots of maximum change from baseline for each of the three groups (Robert C et al, N Engl J Med 2015;372:2521-2532).

Case: Mr MB

  • Presentation: diarrhoea for 2 months; colicky abdominal pain with intermittent obstructive symptoms; 8 kg weight loss; dry cough, worse nocturnally.
  • Investigations: LDH three times the upper limit of normal; biopsy S100 positive; BRAF V600E mutation detected. CT of chest and abdomen or pelvis shows disease, with two abdominal lesions circled.
  • Course: switched to pembrolizumab on progression, with paired coronal abdominal or pelvic CT and paired axial chest CT (lung windows) documenting the response to pembrolizumab.

Brain metastases

  • Risk factors for brain metastases: male gender, mucosal primary, head and neck primary, high risk primary, acral lentiginous or nodular subtype, three or more regional lymph nodes, visceral metastases at diagnosis.
  • Combination ipilimumab plus nivolumab gives a response rate of 30 to 56%.
  • Steroids are a problem in this setting.

Closing point

Knowledge only opens the door: compassion and care are where medicine really happens.

Self-test

  1. List the five categories of systemic anti-cancer therapy used in medical oncology.
  2. Describe the phases of the cell cycle in order, saying what happens in each, and state what regulates the transitions.
  3. Describe the stages of mitosis in order with the defining event of each.
  4. State cancer’s share of deaths in New Zealand and list the five cancer types that account for the majority of cases with their percentages.
  5. Distinguish the trend in the number of cancer deaths in New Zealand from the trend in the death rate per 100,000, 1948 to 2012.
  6. List the mechanisms by which cancer therapies act, with the example given for each.
  7. Describe the generic architecture of a signalling pathway that targeted agents exploit, and say which drug type is directed at which part of it.
  8. Name the main signalling pathways targeted by these agents, and describe the EGFR cascade shown in the lecture including the points at which the slides show blockade.
  9. Which drug opens the TKI approval timeline and in what year, and what do the one, two and three asterisk footnote markers denote?
  10. What does the slide on the original targeted therapy show about Beatson’s 1896 paper?
  11. Outline the sequence of events from 1968 to 1982 that established oestrogen receptor targeting in breast cancer, and state the lesson drawn from it.
  12. What did the 2001 STI571 report in gastrointestinal stromal tumour show, and why is it presented as a landmark?
  13. Explain the monoclonal antibody humanisation spectrum, giving the mouse protein content and suffix at each step, and name the two associated effects.
  14. For rituximab, cetuximab, trastuzumab, panitumumab, bevacizumab and ramucirumab, give the structure, target and indication of each.
  15. Describe the composition of T-DM1 and the five steps by which an antibody-drug conjugate kills a target cell.
  16. List the toxicity categories of targeted therapies with the specific examples given for each.
  17. Describe the three mechanisms of anti-VEGF tyrosine kinase inhibitor induced hypertension, and state which agent produced the largest rise in systolic blood pressure and by how much.
  18. Describe the CRYSTAL trial result, then explain how it changes when patients are stratified by KRAS status.
  19. Why do the KRAS-stratified curves in the FOLFIRI-alone arm matter to the interpretation of KRAS as a biomarker?
  20. Describe the design and primary analysis result of the New EPOC study, and explain what it adds to the CRYSTAL finding.
  21. Describe the rationale for adding panitumumab to chemotherapy in oesophagogastric cancer, and the overall survival result of REAL3.
  22. List the lessons the lecture draws from EGFR targeting.
  23. Distinguish the mechanism of anti-CTLA-4 therapy from that of anti-PD-1 or anti-PD-L1 therapy.
  24. List the inhibitory tumour cell to T cell or NK cell receptor-ligand pairs shown on the checkpoint figure.
  25. List the organ systems affected by immunotherapy toxicity, with at least two examples for each.
  26. Distinguish the four clinical scenarios of primary, adaptive and acquired resistance to immunotherapy.
  27. List the tumour cell intrinsic and tumour cell extrinsic mechanisms of primary and adaptive resistance to immunotherapy.
  28. Describe how melanoma is subdivided molecularly in this lecture, and what the range of clinical photographs is used to make.
  29. Describe the efficacy and toxicity of vemurafenib in V600E mutated melanoma, including its progression-free survival comparison against dacarbazine.
  30. A man presents with two months of diarrhoea, colicky abdominal pain with intermittent obstructive symptoms, 8 kg weight loss and a nocturnal dry cough. LDH is three times the upper limit of normal. What investigation results established the diagnosis and its molecular subtype, and what happened when he progressed?
  31. List the risk factors for brain metastases in melanoma, give the response rate for combination ipilimumab plus nivolumab, and state the treatment problem flagged.
  32. Integrative: using the EGFR and melanoma material, explain why over-expression of a pathway is “a hint not a home run” and what has to be established before a targeted agent is used.

Answers

Graph View

Backlinks