Predicting progression-free survival after systemic therapy in advanced head and neck cancer: Bayesian regression and model development

Recurrent (R) or metastatic (M) head and neck squamous cell carcinoma (HNSCC) is associated with a poor prognosis. For many years, the standard-of-care first-line systemic treatment was the EXTREME regimen, consisting of a platinum-based chemotherapy regimen and cetuximab, an anti-EGFR monoclonal antibody (Van Cutsem et al., 2009). The KEYNOTE-048 trial in 2019 changed the treatment paradigm for these patients by incorporating pembrolizumab, an immune checkpoint inhibitor, into the first-line setting (Burtness et al., 2019). However, the EXTREME regimen remains a first-line standard-of-care for a substantial number of patients, specifically those with programmed death ligand 1 (PD-L1) negative tumours or those with contraindications to the use of anti-PD1 immunotherapy.

While effective, these regimens are associated with significant toxicities. One of the key challenges for the treating physician is to identify the patients who would benefit from either of these treatment regimens. A predictive biomarker signature for patients with advanced HNSCC will help individualize discussions with patients regarding the risk-benefit balance of treatment and may guide patients who are likely to perform poorly towards alternative therapy regimens or clinical trials.

The absence of predictive biomarkers in this patient cohort represents a significant clinical unmet need. Until the development of PD-L1 as a biomarker for immunotherapy, efforts to generate biomarkers in HNSCC have focused on gene expression profiles, which are dependent on the availability of tumour tissue and are only performed on pre-treatment samples (Bossi et al., 2016; You et al., 2019). Signatures based on a single biological modality and taken at a single timepoint may be insufficient to predict outcomes, as response to therapy relies on a dynamic interplay between cancer genomics, immune profile, tumour microenvironment, and clinicopathological characteristics of the patient receiving treatment (Cheerla and Gevaert, 2019; Huang et al., 2019).

Efforts to develop a machine learning model to stratify survival risk by combining genetic and clinicopathological characteristics have revealed some success in advanced oral squamous cell carcinoma (Tseng et al., 2020). We hypothesize that a multimodal analysis, taking into account both clinicopathological and laboratory-based biological covariates at different timepoints, would provide better predictive value. Furthermore, by only including covariates that can be obtained from a blood biopsy, patients could be easily screened for discovered biomarkers.

We prospectively collected peripheral blood samples from a Phase 2 trial in R/M HNSCC (NTC02633800) (Forster et al., 2019), which utilized cetuximab with platinum therapy as a backbone, and conducted a parallel exploratory analysis with the aim of generating a biomarker signature which would predict outcomes to treatment. We hypothesized that the detailed definition of a broad immune cell signature could contribute to the development of assays employing liquid biopsies to predict clinical outcomes. We also incorporated the analysis of two circulating microRNAs (miRNAs) from exosomes: miR-21-5p and miR-142-3p, which have previously demonstrated prognostic and predictive utility (Summerer et al., 2015; Vahabi et al., 2021). As the trial investigated the efficacy of an anti-ErbB3 antibody, patritumab, administered alongside an anti-EGFR antibody, we simultaneously analysed EGFR-ErbB3 dimerization using Förster resonance energy transfer (FRET) and included it in our analysis.

By extracting information from patient samples at baseline and after the first cycle of treatment within this trial, we aimed to generate a multimodal predictive signature for the systemic therapy based on a novel Bayesian multivariate model. The immune populations driving the risk signature were then prospectively validated in an independent cohort of patients from a Phase 2 trial evaluating avelumab in combination with cetuximab in R/M HNSCC (the EACH trial; NCT03494322). Blood samples were collected at baseline and post first dose of cetuximab before administration of avelumab. Validation was performed against the best overall response, as assessed by iRECIST, showing strong correlation between biomarkers identified in the risk signature and therapeutic response despite changes in treatment regimen. We then used imaging mass cytometry driven by the risk signature to identify correlate changes in systemic immune populations with intra-tumoural immune subsets that colocalize at the tumoural level and appear to be key to response. This risk signature can serve as a non-invasive risk stratification for patients with R/M HNSCC using only peripheral blood, guiding the clinician toward the likelihood of success early during the treatment course.

There is a clinical unmet need to identify predictive biomarkers for treatment in head and neck cancer. Gene expression profiling has revealed promising initial results in this domain but have been limited to HPV-positive HNSCC which inherently have better prognoses. The ratio of neutrophils to lymphocytes (NLR) in the peripheral blood of HNSCC patients prior to treatment has been extensively investigated for its ability to predict disease outcome. Two recent meta-analyses on the subject have identified that a high pre-treatment NLR correlated with worse OS with respective HRs of 1.69 (95% CI: 1.47–1.93; p<0.001) and HR of 1.78 (95% CI: 1.53–2.07; p<0.0001) (Mascarella et al., 2018; Takenaka et al., 2018). Using our combined risk signature, we were able to predict disease outcome more accurately across two separate treatment modalities. Other recent biomarker research has focused on predicting response to immune checkpoint blockade by analysing tissue-based biomarkers, such as PD-L1 expression levels, but when used in isolation have not been sufficiently predictive at identifying patients who would benefit (Burtness et al., 2019).

While unimodal biomarkers may offer some predictive value, the biology of HNSCC and likelihood of response to treatment is likely to be dictated by an interplay between tumour immunity, genomic signatures, and a host of clinicopathological characteristics which can only be assessed by a multivariate analysis. There has been an increased interest in peripheral blood-based liquid biopsies in recent years, particularly in the context of PBMC analysis (Nixon et al., 2019). The ability to extract predictive biomarkers from a blood-based liquid biopsy mitigates certain limitations posed by tissue biopsies – particularly the tissue accessibility, technical expertise to obtain the biopsy, patient frailty, and the amount of tissue available. The ease of obtaining liquid biopsies also facilitates longitudinal monitoring of response to treatment. Moreover, PBMCs offer a much more systemic snapshot of the immune response which overcomes intra-tumoural heterogeneity that results in non-uniform distribution of biomarkers. The potential of PBMCs to unravel the tumoural immune landscape is slowly being realized, thanks to advancements in the fields of high-dimensional flow cytometry, single-cell RNA sequencing, and cytometry by time of flight. However, such techniques generate an excessive number of variables which require powerful statistical analysis tools to overcome pitfalls such as overfitting (Shalabi et al., 2018).

To our knowledge, the current study is one of the first examples of integrating multiple biological covariates derived from peripheral blood and patient clinical data to generate a signature which predicts treatment response and that has been validated across multiple therapeutic modalities.

By employing cross-validation iterations to estimate training and validation errors, implementing advanced overfitting correlation protocols, using built-in corrections for informative data missingness, and probabilistic covariate removal, we were able to derive a robust optimal covariate set which correlates with PFS. This combination of analyses has been shown to produce robust signatures that do generalize to uncaptured data (Barber et al., 2020).

The biological components of the predictive model warrant discussion. The only clinical covariate to feature in the combined risk signature was the hypopharyngeal SCC sub-site, which was adversely correlated with PFS. This corroborates previous findings that the 5-year relative survival of patients with hypopharyngeal SCC is consistently the worst amongst different anatomical HNSCC sub-sites (Gatta et al., 2017; Machiels et al., 2020). The propensity of hypopharyngeal tumours to present at the de novo advanced stage (Cadoni et al., 2017) and the density of submucosal lymphatics in this anatomical region translates into these patients inherently performing worse – lending support to the robust nature of our predictive signature. The notable absence of patritumab (denoted as ‘Drug’) in our predictive signatures is also consistent with the outcome of the Phase 2 clinical trial where the addition of this investigational medicinal product did not produce any benefit to PFS (Forster et al., 2019).

Another key discovery in this study is the significance of analysis of sequential samples even at early stages of therapy. This, we believe, is important for the analysis of lab-based biological variables. There is a high level of inter-individual heterogeneity in the relative abundance of PBMC subpopulations even between healthy donors. This inherently weakens the predictive power of covariates based on raw PBMC populations. We observe this phenomenon in our dataset as the combined risk signature had higher predictive power than the baseline one. By opting to present the biological data obtained from the first post-treatment biopsy as change relative to the baseline biopsy, we obtained a much more predictive risk signature in our combined model. This is highlighted in the post-treatment change in CD8CM percentage relative to baseline value which is not only the strongest predictor of PFS in the combined model, but also has a greater univariate predictive value than the whole baseline signature. This validates the potential of our approach has in deconvoluting observed biological phenomena due to patient heterogeneity from that induced by drug treatment, both of which are key to biomarker discovery.

We also validated our risk signature in a second cohort thus allowing us to successfully identify predictive biological biomarkers for therapy across different treatment modalities. The fact that the first trial involved targeted therapies and the second a combination of anti-EGFR and anti-PD-L1 therapy, yet the same blood kinetic change can still distinguish between responders and non-responders is novel and has not been seen before according to our knowledge. The limitation is that the independent validation cohort is small despite the fact that the result we obtained with the present number of validation samples did reach significance (Kandall tau = 0.725, with p-value 0.018). This further highlights the power of multivariate analysis as none of the univariates identified in the test cohort independently predicted treatment outcome in the validation cohort, yet the risk signature strongly correlated with a poorer outcome. We believe that understanding the interplay of key immune subsets between the tumoural microenvironment and the circulation will be key to biomarker development. We focused on two covariates from the risk signature based on their predictive power in both cohorts. Analysis of immune populations at the tissue level allowed us to identify tissue counterparts of peripheral CD33⁺HLADR⁺ monocytes and CD8CM that not only colocalized at the tissue level but also were favourable to a good outcome. Hence, we hypothesize to have uncovered a two-checkpoint system for predicting response to systemic therapy which merits further investigation.

Firstly, a favourable therapeutic outcome requires the presence of high pre-treatment levels of CD14⁺CD16⁺CD33⁺HLADR^high monocytes. These correlated with a similar population of CD74⁺CD68⁺ population of tumoural myeloid cells. Myeloid cells are plastic yet given the similarity in the marker expression patterns between the peripheral CD33⁺CD14⁺CD16⁺HLADR^high monocyte population and the CD14⁺CD33⁺CD74⁺CD68⁺ tissue macrophages, as well as the strong correlation between their corresponding blood and tissue abundance (Figure 6D), we hypothesize that they represent the same group of cells that gain CD68 expression as they enter the tissue. This monocyte population matches the population of intermediate monocytes described in the literature shown to have a key role in antigen presentation (Cravens et al., 2007; Zawada et al., 2011; Wong et al., 2011). Given the correlation between tissue levels CD8⁺ memory T cell, peripheral intermediate monocytes, and the tissue colocalization of the CD8⁺ memory T cells with the CD74⁺ macrophages, we can imagine a role for our myeloid population in inducing post-treatment expansion of CD8CM potentially via a mechanism involving cross-presentation of antigens released by drug-induced cancer cell death (Colbert et al., 2020). Indeed, cetuximab treatment can be a source of tumour-derived antigens as it has been shown to increase TCR diversity in peripheral T cells as well as cross-presentation by antigen-presenting cells (Kansy et al., 2018; Srivastava et al., 2013). Ideally, we would require analysis of the draining lymph node to fully understand how these cells move between the tumour and the circulation and their contribution to the expansion of the CD8CM, a role generally accomplished by dendritic cells at the level of the lymph node. Nevertheless, monocyte-derived macrophages have been shown to activate and expand tissue resident memory T cells (Trm) (Snyder et al., 2021; Low et al., 2020; Muntjewerff et al., 2020; Chu et al., 2020).

An on-treatment increase in the level of circulating CD8CM significantly correlated with a worse disease outcome in patients who received two regimens of systemic treatments within two trials. Interestingly, high levels of pre-treatment CD8CM cells correlating with a favourable treatment outcome in HNSCC has been previously and independently reported by the Whiteside group indirectly validating our results (Czystowska et al., 2013). While an increase in CD8CM correlating with a poorer outcome in both our datasets seems counter-intuitive, it is key to realize that this increase is only seen at the levels of peripheral CD8CM which inversely correlate with tissue Trm (Figure 6D). Using single-cell sequencing and mapping of T cell receptor clonality at the levels of TILs and peripheral T cells, researchers have definitively proven that tumour reactive T cell clones can be found in the periphery (Fairfax et al., 2020; Kok et al., 2020; Pauken et al., 2021; Padrón et al., 2022). The movement of these cells between the tumour and draining lymph node was recently investigated using photoactivable T cells in mice which showed that there is continuous recruitment of memory T cells from the periphery into the tumour which replenishes the pool of tissue Trm identified by CD103 and TCF1 expression. However, this recruitment is usually matched by an egress of Trm cells out of the tumour to the draining lymph nodes which is accompanied by decreased expression CXCR6 (Li et al., 2022). The peripheral origin of Trm cells is still under investigation with recent data suggesting that CD8CM are not only capable of homing into the inflamed tissue but also undergoing transcriptional reprogramming to a stable Trm phenotype (Matos et al., 2022). We hypothesize that the predictive power of the population of peripheral CD8CM stems from its ability to home to the tumour and contribute to the Trm population which has recently been identified as key player in anti-tumour immunity in HNSCC as well as other cancers mainly in the context of immunotherapy but also in chemo/radiotherapy (Han et al., 2021; Ida et al., 2021; Savas et al., 2018; Djenidi et al., 2015; Luoma et al., 2022). Unfortunately, one key shortcoming of our study is the lack of Trm-specific markers such as CD103, CD69, and TCF1 from the imaging mass cytometry panel of antibodies which we believe contributed to this anti-correlation not being stronger.

The accumulation of clonally expanded memory T cells in the periphery in poor responders is the potential result of tumour-induced T cell exclusion which we can infer from the negative correlation observed between peripheral and tumour infiltrating memory T cells (Jerby-Arnon et al., 2018). In this model CD8⁺ memory T cells are being generated only in patients having a large pre-treatment pool of competent APCs yet these anti-tumour CD8⁺ memory cells cannot enter the tumour from the periphery hence accumulating in circulation. Our group is now focused on understanding the different aspects of this model mainly with regards to the specific tumour exclusion of CD8CM T cells.

Another limitation of our work to discuss is the absence of overall survival (OS) data within our current dataset. It would have been interesting to assess whether the immune markers predict survival in the longer term. However, the accuracy of the predictive signature for OS is often diluted by a variety of subsequent treatment regimens.

The present study shows that the combination of biomarkers established prospectively by liquid biopsies early in the treatment course offers potential for the provision of personalized treatments to patients (Nenclares et al., 2021). The risk signature drove the discovery of synergies between the systemic and tumoural immune response. Identifying such complex biological interplay as having a role in disease outcome is impossible without multivariate analysis. The post-stratification survival curves in our study demonstrate markedly different PFS outcomes as a testament to this robust statistical model and could represent an invaluable guide to clinicians during the initial stages of treatment.

The data generated in this study and used for multivariate modelling are available from the UCL repository: https://doi.org/10.5522/04/16566207.v1.

1. 1. Barber PR
   2. Flores-Borja F
   3. Alfano G
   4. Ng K
   5. Weitsman G
   6. Dolcetti L
   7. Mustapha R
   8. Wong F
   9. Vicencio JM
   10. Galazi M
   11. Opzoomer J
   12. Arnold J
   13. Kordasti S
   14. Doyle J
   15. Greenberg J
   16. Dillon M
   17. Harrington K
   18. Forster MF
   19. Coolen T
   20. Ng T

   (2021) UCL Research Data Repository

   Head and Neck Cancer Multivariate Blood Data.

   https://doi.org/10.5522/04/16566207.v1

1. 1. Barber PR
   2. Ameer-Beg SM
   3. Gilbey J
   4. Carlin LM
   5. Keppler M
   6. Ng TC
   7. Vojnovic B

   (2009) Multiphoton time-domain fluorescence lifetime imaging microscopy: practical application to protein–protein interactions using global analysis

   Journal of The Royal Society Interface 6:S93–S105.

   https://doi.org/10.1098/rsif.2008.0451.focus

   • Google Scholar
2. 1. Barber PR
   2. Tullis IDC
   3. Pierce GP
   4. Newman RG
   5. Prentice J
   6. Rowley MI
   7. Matthews DR
   8. Ameer-Beg SM
   9. Vojnovic B

   (2013) The gray Institute “ open ” high-content, fluorescence lifetime microscopes

   Journal of Microscopy 251:154–167.

   https://doi.org/10.1111/jmi.12057

   • PubMed
   • Google Scholar
3. 1. Barber PR
   2. Weitsman G
   3. Lawler K
   4. Barrett JE
   5. Rowley M
   6. Rodriguez-Justo M
   7. Fisher D
   8. Gao F
   9. Tullis IDC
   10. Deng J
   11. Brown L
   12. Kaplan R
   13. Hochhauser D
   14. Adams R
   15. Maughan TS
   16. Vojnovic B
   17. Coolen ACC
   18. Ng T

   (2020) Her2-Her3 heterodimer quantification by FRET-FLIM and patient subclass analysis of the coin colorectal trial

   Journal of the National Cancer Institute 112:944–954.

   https://doi.org/10.1093/jnci/djz231

   • PubMed
   • Google Scholar
4. 1. Bossi P
   2. Bergamini C
   3. Siano M
   4. Cossu Rocca M
   5. Sponghini AP
   6. Favales F
   7. Giannoccaro M
   8. Marchesi E
   9. Cortelazzi B
   10. Perrone F
   11. Pilotti S
   12. Locati LD
   13. Licitra L
   14. Canevari S
   15. De Cecco L

   (2016) Functional genomics uncover the biology behind the responsiveness of head and neck squamous cell cancer patients to cetuximab

   Clinical Cancer Research 22:3961–3970.

   https://doi.org/10.1158/1078-0432.CCR-15-2547

   • PubMed
   • Google Scholar
5. 1. Bronte V
   2. Brandau S
   3. Chen SH
   4. Colombo MP
   5. Frey AB
   6. Greten TF
   7. Mandruzzato S
   8. Murray PJ
   9. Ochoa A
   10. Ostrand-Rosenberg S
   11. Rodriguez PC
   12. Sica A
   13. Umansky V
   14. Vonderheide RH
   15. Gabrilovich DI

   (2016) Recommendations for myeloid-derived suppressor cell nomenclature and characterization standards

   Nature Communications 7:12150.

   https://doi.org/10.1038/ncomms12150

   • PubMed
   • Google Scholar
6. 1. Burtness B
   2. Harrington KJ
   3. Greil R
   4. Soulières D
   5. Tahara M
   6. de Castro G
   7. Psyrri A
   8. Basté N
   9. Neupane P
   10. Bratland Å
   11. Fuereder T
   12. Hughes BGM
   13. Mesía R
   14. Ngamphaiboon N
   15. Rordorf T
   16. Wan Ishak WZ
   17. Hong RL
   18. González Mendoza R
   19. Roy A
   20. Zhang Y
   21. Gumuscu B
   22. Cheng JD
   23. Jin F
   24. Rischin D
   25. KEYNOTE-048 Investigators

   (2019) Pembrolizumab alone or with chemotherapy versus cetuximab with chemotherapy for recurrent or metastatic squamous cell carcinoma of the head and neck (KEYNOTE-048): a randomised, open-label, phase 3 study

   Lancet 394:1915–1928.

   https://doi.org/10.1016/S0140-6736(19)32591-7

   • PubMed
   • Google Scholar
7. 1. Cadoni G
   2. Giraldi L
   3. Petrelli L
   4. Pandolfini M
   5. Giuliani M
   6. Paludetti G
   7. Pastorino R
   8. Leoncini E
   9. Arzani D
   10. Almadori G
   11. Boccia S

   (2017) Prognostic factors in head and neck cancer: a 10-year retrospective analysis in a single-institution in Italy

   Acta Otorhinolaryngologica Italica 37:458–466.

   https://doi.org/10.14639/0392-100X-1246

   • PubMed
   • Google Scholar
8. 1. Cheerla A
   2. Gevaert O

   (2019) Deep learning with multimodal representation for pancancer prognosis prediction

   Bioinformatics 35:i446–i454.

   https://doi.org/10.1093/bioinformatics/btz342

   • PubMed
   • Google Scholar
9. 1. Chu KL
   2. Batista NV
   3. Girard M
   4. Watts TH

   (2020) Monocyte-derived cells in tissue-resident memory T cell formation

   Journal of Immunology 204:477–485.

   https://doi.org/10.4049/jimmunol.1901046

   • PubMed
   • Google Scholar
10. 1. Colbert JD
    2. Cruz FM
    3. Rock KL

    (2020) Cross-Presentation of exogenous antigens on MHC I molecules

    Current Opinion in Immunology 64:1–8.

    https://doi.org/10.1016/j.coi.2019.12.005

    • PubMed
    • Google Scholar
11. 1. Cravens PD
    2. Hayashida K
    3. Davis LS
    4. Nanki T
    5. Lipsky PE

    (2007) Human peripheral blood dendritic cells and monocyte subsets display similar chemokine receptor expression profiles with differential migratory responses

    Scandinavian Journal of Immunology 65:514–524.

    https://doi.org/10.1111/j.1365-3083.2007.01933.x

    • PubMed
    • Google Scholar
12. 1. Czystowska M
    2. Gooding W
    3. Szczepanski MJ
    4. Lopez-Abaitero A
    5. Ferris RL
    6. Johnson JT
    7. Whiteside TL

    (2013) The immune signature of CD8 (+) CCR7 (+) T cells in the peripheral circulation associates with disease recurrence in patients with HNSCC

    Clinical Cancer Research 19:889–899.

    https://doi.org/10.1158/1078-0432.CCR-12-2191

    • PubMed
    • Google Scholar
13. 1. Djenidi F
    2. Adam J
    3. Goubar A
    4. Durgeau A
    5. Meurice G
    6. de Montpréville V
    7. Validire P
    8. Besse B
    9. Mami-Chouaib F

    (2015) CD8+CD103+ tumor-infiltrating lymphocytes are tumor-specific tissue-resident memory T cells and a prognostic factor for survival in lung cancer patients

    Journal of Immunology 194:3475–3486.

    https://doi.org/10.4049/jimmunol.1402711

    • PubMed
    • Google Scholar
14. 1. Fairfax BP
    2. Taylor CA
    3. Watson RA
    4. Nassiri I
    5. Danielli S
    6. Fang H
    7. Mahé EA
    8. Cooper R
    9. Woodcock V
    10. Traill Z
    11. Al-Mossawi MH
    12. Knight JC
    13. Klenerman P
    14. Payne M
    15. Middleton MR

    (2020) Peripheral CD8+ T cell characteristics associated with durable responses to immune checkpoint blockade in patients with metastatic melanoma

    Nature Medicine 26:193–199.

    https://doi.org/10.1038/s41591-019-0734-6

    • PubMed
    • Google Scholar
15. 1. Forster MD
    2. Dillon MT
    3. Kocsis J
    4. Remenár É
    5. Pajkos G
    6. Rolland F
    7. Greenberg J
    8. Harrington KJ

    (2019) Patritumab or placebo, with cetuximab plus platinum therapy in recurrent or metastatic squamous cell carcinoma of the head and neck: a randomised phase II study

    European Journal of Cancer 123:36–47.

    https://doi.org/10.1016/j.ejca.2019.08.017

    • PubMed
    • Google Scholar
16. 1. Gatta G
    2. Capocaccia R
    3. Botta L
    4. Mallone S
    5. De Angelis R
    6. Ardanaz E
    7. Comber H
    8. Dimitrova N
    9. Leinonen MK
    10. Siesling S
    11. van der Zwan JM
    12. Van Eycken L
    13. Visser O
    14. Žakelj MP
    15. Anderson LA
    16. Bella F
    17. Kaire I
    18. Otter R
    19. Stiller CA
    20. Trama A
    21. RARECAREnet working group

    (2017) Burden and centralised treatment in Europe of rare tumours: results of rarecarenet-a population-based study

    The Lancet. Oncology 18:1022–1039.

    https://doi.org/10.1016/S1470-2045(17)30445-X

    • PubMed
    • Google Scholar
17. 1. Grigoriadis A
    2. Gazinska P
    3. Pai T
    4. Irhsad S
    5. Wu Y
    6. Millis R
    7. Naidoo K
    8. Owen J
    9. Gillett CE
    10. Tutt A
    11. Coolen AC
    12. Pinder SE

    (2018) Histological scoring of immune and stromal features in breast and axillary lymph nodes is prognostic for distant metastasis in lymph node-positive breast cancers

    The Journal of Pathology. Clinical Research 4:39–54.

    https://doi.org/10.1002/cjp2.87

    • PubMed
    • Google Scholar
18. 1. Han J
    2. Zhao Y
    3. Shirai K
    4. Molodtsov A
    5. Kolling FW
    6. Fisher JL
    7. Zhang P
    8. Yan S
    9. Searles TG
    10. Bader JM
    11. Gui J
    12. Cheng C
    13. Ernstoff MS
    14. Turk MJ
    15. Angeles CV

    (2021) Resident and circulating memory T cells persist for years in melanoma patients with durable responses to immunotherapy

    Nature Cancer 2:300–311.

    https://doi.org/10.1038/s43018-021-00180-1

    • PubMed
    • Google Scholar
19. 1. Harrell FE
    2. Lee KL
    3. Califf RM
    4. Pryor DB
    5. Rosati RA

    (1984) Regression modelling strategies for improved prognostic prediction

    Statistics in Medicine 3:143–152.

    https://doi.org/10.1002/sim.4780030207

    • PubMed
    • Google Scholar
20. 1. Huang Z
    2. Zhan X
    3. Xiang S
    4. Johnson TS
    5. Helm B
    6. Yu CY
    7. Zhang J
    8. Salama P
    9. Rizkalla M
    10. Han Z
    11. Huang K

    (2019) Salmon: survival analysis learning with multi-omics neural networks on breast cancer

    Frontiers in Genetics 10:166.

    https://doi.org/10.3389/fgene.2019.00166

    • PubMed
    • Google Scholar
21. 1. Ida S
    2. Takahashi H
    3. Kawabata-Iwakawa R
    4. Mito I
    5. Tada H
    6. Chikamatsu K

    (2021) Tissue-resident memory T cells correlate with the inflammatory tumor microenvironment and improved prognosis in head and neck squamous cell carcinoma

    Oral Oncology 122:105508.

    https://doi.org/10.1016/j.oraloncology.2021.105508

    • PubMed
    • Google Scholar
22. 1. Jerby-Arnon L
    2. Shah P
    3. Cuoco MS
    4. Rodman C
    5. Su MJ
    6. Melms JC
    7. Leeson R
    8. Kanodia A
    9. Mei S
    10. Lin JR
    11. Wang S
    12. Rabasha B
    13. Liu D
    14. Zhang G
    15. Margolais C
    16. Ashenberg O
    17. Ott PA
    18. Buchbinder EI
    19. Haq R
    20. Hodi FS
    21. Boland GM
    22. Sullivan RJ
    23. Frederick DT
    24. Miao B
    25. Moll T
    26. Flaherty KT
    27. Herlyn M
    28. Jenkins RW
    29. Thummalapalli R
    30. Kowalczyk MS
    31. Cañadas I
    32. Schilling B
    33. Cartwright ANR
    34. Luoma AM
    35. Malu S
    36. Hwu P
    37. Bernatchez C
    38. Forget MA
    39. Barbie DA
    40. Shalek AK
    41. Tirosh I
    42. Sorger PK
    43. Wucherpfennig K
    44. Van Allen EM
    45. Schadendorf D
    46. Johnson BE
    47. Rotem A
    48. Rozenblatt-Rosen O
    49. Garraway LA
    50. Yoon CH
    51. Izar B
    52. Regev A

    (2018) A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade

    Cell 175:984–997.

    https://doi.org/10.1016/j.cell.2018.09.006

    • PubMed
    • Google Scholar
23. 1. Kansy BA
    2. Shayan G
    3. Jie H-B
    4. Gibson SP
    5. Lei YL
    6. Brandau S
    7. Lang S
    8. Schmitt NC
    9. Ding F
    10. Lin Y
    11. Ferris RL

    (2018) T cell receptor richness in peripheral blood increases after cetuximab therapy and correlates with therapeutic response

    Oncoimmunology 7:e1494112.

    https://doi.org/10.1080/2162402X.2018.1494112

    • PubMed
    • Google Scholar
24. 1. Kok L
    2. Dijkgraaf FE
    3. Urbanus J
    4. Bresser K
    5. Vredevoogd DW
    6. Cardoso RF
    7. Perié L
    8. Beltman JB
    9. Schumacher TN

    (2020) A committed tissue-resident memory T cell precursor within the circulating CD8+ effector T cell pool

    The Journal of Experimental Medicine 217:10.

    https://doi.org/10.1084/jem.20191711

    • PubMed
    • Google Scholar
25. 1. Li Z
    2. Tuong ZK
    3. Dean I
    4. Willis C
    5. Gaspal F
    6. Fiancette R
    7. Idris S
    8. Kennedy B
    9. Ferdinand JR
    10. Peñalver A
    11. Cabantous M
    12. Murtuza Baker S
    13. Fry JW
    14. Carlesso G
    15. Hammond SA
    16. Dovedi SJ
    17. Hepworth MR
    18. Clatworthy MR
    19. Withers DR

    (2022) In vivo labeling reveals continuous trafficking of TCF-1+ T cells between tumor and lymphoid tissue

    The Journal of Experimental Medicine 219:e20210749.

    https://doi.org/10.1084/jem.20210749

    • PubMed
    • Google Scholar
26. 1. Low JS
    2. Farsakoglu Y
    3. Amezcua Vesely MC
    4. Sefik E
    5. Kelly JB
    6. Harman CCD
    7. Jackson R
    8. Shyer JA
    9. Jiang X
    10. Cauley LS
    11. Flavell RA
    12. Kaech SM

    (2020) Tissue-Resident memory T cell reactivation by diverse antigen-presenting cells imparts distinct functional responses

    The Journal of Experimental Medicine 217:e20192291.

    https://doi.org/10.1084/jem.20192291

    • PubMed
    • Google Scholar
27. 1. Luoma AM
    2. Suo S
    3. Wang Y
    4. Gunasti L
    5. Porter CBM
    6. Nabilsi N
    7. Tadros J
    8. Ferretti AP
    9. Liao S
    10. Gurer C
    11. Chen Y-H
    12. Criscitiello S
    13. Ricker CA
    14. Dionne D
    15. Rozenblatt-Rosen O
    16. Uppaluri R
    17. Haddad RI
    18. Ashenberg O
    19. Regev A
    20. Van Allen EM
    21. MacBeath G
    22. Schoenfeld JD
    23. Wucherpfennig KW

    (2022) Tissue-Resident memory and circulating T cells are early responders to pre-surgical cancer immunotherapy

    Cell 185:2918–2935.

    https://doi.org/10.1016/j.cell.2022.06.018

    • PubMed
    • Google Scholar
28. 1. Machiels J-P
    2. René Leemans C
    3. Golusinski W
    4. Grau C
    5. Licitra L
    6. Gregoire V
    7. EHNS Executive Board. Electronic address: secretariat@ehns.org
    8. ESMO Guidelines Committee. Electronic address: clinicalguidelines@esmo.org
    9. ESTRO Executive Board. Electronic address: info@estro.org

    (2020) Squamous cell carcinoma of the oral cavity, larynx, oropharynx and hypopharynx: EHNS-ESMO-ESTRO clinical practice guidelines for diagnosis, treatment and follow-up

    Annals of Oncology 31:1462–1475.

    https://doi.org/10.1016/j.annonc.2020.07.011

    • PubMed
    • Google Scholar
29. 1. Mascarella MA
    2. Mannard E
    3. Silva SD
    4. Zeitouni A

    (2018) Neutrophil-To-Lymphocyte ratio in head and neck cancer prognosis: a systematic review and meta-analysis

    Head & Neck 40:1091–1100.

    https://doi.org/10.1002/hed.25075

    • PubMed
    • Google Scholar
30. 1. Matos TR
    2. Gehad A
    3. Teague JE
    4. Dyring-Andersen B
    5. Benezeder T
    6. Dowlatshahi M
    7. Crouch J
    8. Watanabe Y
    9. O’Malley JT
    10. Kupper TS
    11. Yang C
    12. Watanabe R
    13. Clark RA

    (2022) Central memory T cells are the most effective precursors of resident memory T cells in human skin

    Science Immunology 7:eabn1889.

    https://doi.org/10.1126/sciimmunol.abn1889

    • PubMed
    • Google Scholar
31. 1. Monypenny J
    2. Milewicz H
    3. Flores-Borja F
    4. Weitsman G
    5. Cheung A
    6. Chowdhury R
    7. Burgoyne T
    8. Arulappu A
    9. Lawler K
    10. Barber PR
    11. Vicencio JM
    12. Keppler M
    13. Wulaningsih W
    14. Davidson SM
    15. Fraternali F
    16. Woodman N
    17. Turmaine M
    18. Gillett C
    19. Franz D
    20. Quezada SA
    21. Futter CE
    22. Von Kriegsheim A
    23. Kolch W
    24. Vojnovic B
    25. Carlton JG
    26. Ng T

    (2018) Alix regulates tumor-mediated immunosuppression by controlling EGFR activity and PD-L1 presentation

    Cell Reports 24:630–641.

    https://doi.org/10.1016/j.celrep.2018.06.066

    • PubMed
    • Google Scholar
32. 1. Muntjewerff EM
    2. Meesters LD
    3. van den Bogaart G

    (2020) Antigen cross-presentation by macrophages

    Frontiers in Immunology 11:1276.

    https://doi.org/10.3389/fimmu.2020.01276

    • PubMed
    • Google Scholar
33. 1. Nenclares P
    2. Gunn L
    3. Soliman H
    4. Bover M
    5. Trinh A
    6. Leslie I
    7. Wong KH
    8. Melcher A
    9. Newbold K
    10. Nutting CM
    11. Ap Dafydd D
    12. Bhide SA
    13. Harrington K

    (2021) On-Treatment immune prognostic score for patients with relapsed and/or metastatic head and neck squamous cell carcinoma treated with immunotherapy

    Journal for Immunotherapy of Cancer 9:e002718.

    https://doi.org/10.1136/jitc-2021-002718

    • PubMed
    • Google Scholar
34. 1. Nixon AB
    2. Schalper KA
    3. Jacobs I
    4. Potluri S
    5. Wang IM
    6. Fleener C

    (2019) Peripheral immune-based biomarkers in cancer immunotherapy: can we realize their predictive potential?

    Journal for Immunotherapy of Cancer 7:325.

    https://doi.org/10.1186/s40425-019-0799-2

    • PubMed
    • Google Scholar
35. 1. Padrón LJ
    2. Maurer DM
    3. O’Hara MH
    4. O’Reilly EM
    5. Wolff RA
    6. Wainberg ZA
    7. Ko AH
    8. Fisher G
    9. Rahma O
    10. Lyman JP
    11. Cabanski CR
    12. Yu JX
    13. Pfeiffer SM
    14. Spasic M
    15. Xu J
    16. Gherardini PF
    17. Karakunnel J
    18. Mick R
    19. Alanio C
    20. Byrne KT
    21. Hollmann TJ
    22. Moore JS
    23. Jones DD
    24. Tognetti M
    25. Chen RO
    26. Yang X
    27. Salvador L
    28. Wherry EJ
    29. Dugan U
    30. O’Donnell-Tormey J
    31. Butterfield LH
    32. Hubbard-Lucey VM
    33. Ibrahim R
    34. Fairchild J
    35. Bucktrout S
    36. LaVallee TM
    37. Vonderheide RH

    (2022) Sotigalimab and/or nivolumab with chemotherapy in first-line metastatic pancreatic cancer: clinical and immunologic analyses from the randomized phase 2 Prince trial

    Nature Medicine 28:1167–1177.

    https://doi.org/10.1038/s41591-022-01829-9

    • PubMed
    • Google Scholar
36. 1. Pauken KE
    2. Shahid O
    3. Lagattuta KA
    4. Mahuron KM
    5. Luber JM
    6. Lowe MM
    7. Huang L
    8. Delaney C
    9. Long JM
    10. Fung ME
    11. Newcomer K
    12. Tsai KK
    13. Chow M
    14. Guinn S
    15. Kuchroo JR
    16. Burke KP
    17. Schenkel JM
    18. Rosenblum MD
    19. Daud AI
    20. Sharpe AH
    21. Singer M

    (2021) Single-Cell analyses identify circulating anti-tumor CD8 T cells and markers for their enrichment

    The Journal of Experimental Medicine 218:e20200920.

    https://doi.org/10.1084/jem.20200920

    • PubMed
    • Google Scholar
37. 1. Rowley MI
    2. Coolen ACC
    3. Vojnovic B
    4. Barber PR

    (2016) Robust Bayesian fluorescence lifetime estimation, decay model selection and instrument response determination for low-intensity FLIM imaging

    PLOS ONE 11:e0158404.

    https://doi.org/10.1371/journal.pone.0158404

    • PubMed
    • Google Scholar
38. 1. Savas P
    2. Virassamy B
    3. Ye C
    4. Salim A
    5. Mintoff CP
    6. Caramia F
    7. Salgado R
    8. Byrne DJ
    9. Teo ZL
    10. Dushyanthen S
    11. Byrne A
    12. Wein L
    13. Luen SJ
    14. Poliness C
    15. Nightingale SS
    16. Skandarajah AS
    17. Gyorki DE
    18. Thornton CM
    19. Beavis PA
    20. Fox SB
    21. Darcy PK
    22. Speed TP
    23. Mackay LK
    24. Neeson PJ
    25. Loi S
    26. Kathleen Cuningham Foundation Consortium for Research into Familial Breast Cancer

    (2018) Single-cell profiling of breast cancer T cells reveals a tissue-resident memory subset associated with improved prognosis

    Nature Medicine 24:986–993.

    https://doi.org/10.1038/s41591-018-0078-7

    • PubMed
    • Google Scholar
39. 1. Shalabi A
    2. Inoue M
    3. Watkins J
    4. De Rinaldis E
    5. Coolen AC

    (2018) Bayesian clinical classification from high-dimensional data: signatures versus variability

    Statistical Methods in Medical Research 27:336–351.

    https://doi.org/10.1177/0962280216628901

    • PubMed
    • Google Scholar
40. 1. Snyder ME
    2. Sembrat J
    3. Noda K
    4. Myerburg MM
    5. Craig A
    6. Mitash N
    7. Harano T
    8. Furukawa M
    9. Pilewski J
    10. McDyer J
    11. Rojas M
    12. Sanchez P

    (2021) Human lung-resident macrophages colocalize with and provide costimulation to PD1hi tissue-resident memory T cells

    American Journal of Respiratory and Critical Care Medicine 203:1230–1244.

    https://doi.org/10.1164/rccm.202006-2403OC

    • PubMed
    • Google Scholar
41. 1. Srivastava RM
    2. Lee SC
    3. Andrade Filho PA
    4. Lord CA
    5. Jie H-B
    6. Davidson HC
    7. López-Albaitero A
    8. Gibson SP
    9. Gooding WE
    10. Ferrone S
    11. Ferris RL

    (2013) Cetuximab-activated natural killer and dendritic cells collaborate to trigger tumor antigen-specific T-cell immunity in head and neck cancer patients

    Clinical Cancer Research 19:1858–1872.

    https://doi.org/10.1158/1078-0432.CCR-12-2426

    • PubMed
    • Google Scholar
42. 1. Summerer I
    2. Unger K
    3. Braselmann H
    4. Schuettrumpf L
    5. Maihoefer C
    6. Baumeister P
    7. Kirchner T
    8. Niyazi M
    9. Sage E
    10. Specht HM
    11. Multhoff G
    12. Moertl S
    13. Belka C
    14. Zitzelsberger H

    (2015) Circulating microRNAs as prognostic therapy biomarkers in head and neck cancer patients

    British Journal of Cancer 113:76–82.

    https://doi.org/10.1038/bjc.2015.111

    • PubMed
    • Google Scholar
43. 1. Takenaka Y
    2. Oya R
    3. Kitamiura T
    4. Ashida N
    5. Shimizu K
    6. Takemura K
    7. Yamamoto Y
    8. Uno A

    (2018) Prognostic role of neutrophil-to-lymphocyte ratio in head and neck cancer: a meta-analysis

    Head & Neck 40:647–655.

    https://doi.org/10.1002/hed.24986

    • PubMed
    • Google Scholar
44. 1. Tseng YJ
    2. Wang HY
    3. Lin TW
    4. Lu JJ
    5. Hsieh CH
    6. Liao CT

    (2020) Development of a machine learning model for survival risk stratification of patients with advanced oral cancer

    JAMA Network Open 3:e2011768.

    https://doi.org/10.1001/jamanetworkopen.2020.11768

    • PubMed
    • Google Scholar
45. 1. Vahabi M
    2. Blandino G
    3. Di Agostino S

    (2021) Micrornas in head and neck squamous cell carcinoma: a possible challenge as biomarkers, determinants for the choice of therapy and targets for personalized molecular therapies

    Translational Cancer Research 10:3090–3110.

    https://doi.org/10.21037/tcr-20-2530

    • PubMed
    • Google Scholar
46. 1. Van Cutsem E
    2. Köhne C-H
    3. Hitre E
    4. Zaluski J
    5. Chang Chien C-R
    6. Makhson A
    7. D’Haens G
    8. Pintér T
    9. Lim R
    10. Bodoky G
    11. Roh JK
    12. Folprecht G
    13. Ruff P
    14. Stroh C
    15. Tejpar S
    16. Schlichting M
    17. Nippgen J
    18. Rougier P

    (2009) Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer

    The New England Journal of Medicine 360:1408–1417.

    https://doi.org/10.1056/NEJMoa0805019

    • PubMed
    • Google Scholar
47. 1. Wong KL
    2. Tai JJY
    3. Wong WC
    4. Han H
    5. Sem X
    6. Yeap WH
    7. Kourilsky P
    8. Wong SC

    (2011) Gene expression profiling reveals the defining features of the classical, intermediate, and nonclassical human monocyte subsets

    Blood 118:16–31.

    https://doi.org/10.1182/blood-2010-12-326355

    • PubMed
    • Google Scholar
48. 1. You GR
    2. Cheng AJ
    3. Lee LY
    4. Huang YC
    5. Liu H
    6. Chen YJ
    7. Chang JT

    (2019) Prognostic signature associated with radioresistance in head and neck cancer via transcriptomic and bioinformatic analyses

    BMC Cancer 19:64.

    https://doi.org/10.1186/s12885-018-5243-3

    • PubMed
    • Google Scholar
49. 1. Zawada AM
    2. Rogacev KS
    3. Rotter B
    4. Winter P
    5. Marell RR
    6. Fliser D
    7. Heine GH

    (2011) SuperSAGE evidence for CD14++CD16+ monocytes as a third monocyte subset

    Blood 118:50–61.

    https://doi.org/10.1182/blood-2011-01-326827

    • PubMed
    • Google Scholar

Author details

1. Paul R Barber

   1. UCL Cancer Institute, Paul O'Gorman Building, University College London, London, United Kingdom
   2. Comprehensive Cancer Centre, School of Cancer & Pharmaceutical Sciences, King’s College London, London, United Kingdom

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review and editing

   Contributed equally with

   Rami Mustapha, Fabian Flores-Borja, Giovanna Alfano and Kenrick Ng

   Competing interests

   is a shareholder of Nano Clinical Ltd

   "This ORCID iD identifies the author of this article:" 0000-0002-8595-1141

2. Rami Mustapha

   Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom

   Contribution

   Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review and editing

   Contributed equally with

   Paul R Barber, Fabian Flores-Borja, Giovanna Alfano and Kenrick Ng

   Competing interests

   No competing interests declared

3. Fabian Flores-Borja

   Breast Cancer Now Research Unit, School of Cancer & Pharmaceutical Sciences, King’s College London, London, United Kingdom

   Present address

   Centre for Oral Immunobiology and Regenerative Medicine, Barts & The London School of Medicine and Dentistry, Queen Mary University of London, London, United Kingdom

   Contribution

   Data curation, Formal analysis, Investigation, Methodology

   Contributed equally with

   Paul R Barber, Rami Mustapha, Giovanna Alfano and Kenrick Ng

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0881-8822

4. Giovanna Alfano

   Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom

   Contribution

   Data curation, Formal analysis, Investigation

   Contributed equally with

   Paul R Barber, Rami Mustapha, Fabian Flores-Borja and Kenrick Ng

   Competing interests

   No competing interests declared

5. Kenrick Ng

   UCL Cancer Institute, Paul O'Gorman Building, University College London, London, United Kingdom

   Contribution

   Formal analysis, Methodology, Writing – original draft, Writing – review and editing

   Contributed equally with

   Paul R Barber, Rami Mustapha, Fabian Flores-Borja and Giovanna Alfano

   Competing interests

   has received honoraria from Pfizer, GSK/Tesaro and Boheringer Ingleheim, and has had travel/accommodation/expenses paid for by Tesaro

6. Gregory Weitsman

   Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom

   Contribution

   Formal analysis, Investigation

   Competing interests

   No competing interests declared

7. Luigi Dolcetti

   Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom

   Contribution

   Formal analysis, Investigation

   Competing interests

   No competing interests declared

8. Ali Abdulnabi Suwaidan

   Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom

   Contribution

   Writing – review and editing

   Competing interests

   No competing interests declared

9. Felix Wong

   Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom

   Contribution

   Formal analysis, Investigation

   Competing interests

   No competing interests declared

10. Jose M Vicencio

    1. UCL Cancer Institute, Paul O'Gorman Building, University College London, London, United Kingdom
    2. Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom

    Contribution

    Formal analysis, Investigation, Methodology

    Competing interests

    No competing interests declared

11. Myria Galazi

    UCL Cancer Institute, Paul O'Gorman Building, University College London, London, United Kingdom

    Contribution

    Data curation, Investigation

    Competing interests

    No competing interests declared

12. James W Opzoomer

    Tumor Immunology Group, School of Cancer & Pharmaceutical Sciences, King’s College London, London, United Kingdom

    Contribution

    Data curation, Formal analysis, Investigation

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0001-6842-756X

13. James N Arnold

    Tumor Immunology Group, School of Cancer & Pharmaceutical Sciences, King’s College London, London, United Kingdom

    Contribution

    Supervision, Writing – review and editing

    Competing interests

    No competing interests declared

14. Selvam Thavaraj

    Centre for Clinical, Oral & Translational Science, King’s College London, London, United Kingdom

    Contribution

    Methodology

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0001-5720-7422

15. Shahram Kordasti

    Systems Cancer Immunology, School of Cancer & Pharmaceutical Sciences, King’s College London, London, United Kingdom

    Contribution

    Formal analysis, Supervision

    Competing interests

    has received research funding in the form of a grant from Novartis and Celgene

    "This ORCID iD identifies the author of this article:" 0000-0002-0347-4207

16. Jana Doyle

    Daiichi Sankyo Incorporated, Newark, United States

    Contribution

    Resources, Supervision, Funding acquisition, Project administration

    Competing interests

    is in employment with Daichii Sankyo, and has stock and other ownership interests, research funding within Daichii Sankyo and has had travel/accommodation/expenses paid for by Daichii Sankyo

17. Jon Greenberg

    Daiichi Sankyo Incorporated, Newark, United States

    Contribution

    Resources, Funding acquisition, Project administration, Writing – review and editing

    Competing interests

    is in employment with Daichii Sankyo, and has stock and other ownership interests, research funding within Daichii Sankyo and has had travel/accommodation/expenses paid for by Daichii Sankyo

18. Magnus T Dillon

    The Institute of Cancer Research, London, United Kingdom

    Contribution

    Methodology, Project administration

    Competing interests

    No competing interests declared

19. Kevin J Harrington

    The Institute of Cancer Research, London, United Kingdom

    Contribution

    Supervision, Methodology, Project administration, Writing – review and editing

    Competing interests

    has received honoraria from Amgen; Arch Oncology; AstraZeneca; Boehringer-Ingelheim; Bristol-Myers Squibb; Codiak; Inzen; Merck; MSD; Pfizer; Replimune and is on a speakers' bureau for Amgen, AstraZeneca; Bristol-Myers Squibb; Merck, MSD; Pfizer. KH has also received research funding from AstraZeneca, Boehringer-Ingelheim, MSD and Replimune

20. Martin Forster

    UCL Cancer Institute, Paul O'Gorman Building, University College London, London, United Kingdom

    Contribution

    Conceptualization, Supervision, Methodology, Writing – review and editing

    Competing interests

    has received institutional research funding from AstraZeneca, Boehringer-Ingelheim, Merck and MSD and serves in a consulting or advisory role to Achilles, Astrazeneca, Bayer, Bristol-Myers Squibb, Celgene, Guardant Health, Merck, MSD, Nanobiotix, Novartis, Oxford VacMedix, Pfizer, Roche, Takeda, UltraHuman

21. Anthony CC Coolen

    1. Institute for Mathematical and Molecular Biomedicine, King’s College London, London, United Kingdom
    2. Saddle Point Science Ltd, London, United Kingdom

    Contribution

    Conceptualization, Data curation, Formal analysis, Supervision, Methodology

    Competing interests

    has stock and other ownership interests with Saddle Point Science Limited

    "This ORCID iD identifies the author of this article:" 0000-0002-6976-5875

22. Tony Ng

    1. UCL Cancer Institute, Paul O'Gorman Building, University College London, London, United Kingdom
    2. Richard Dimbleby Laboratory of Cancer Research, School of Cancer & Pharmaceutical Sciences, King's College London, London, United Kingdom
    3. Breast Cancer Now Research Unit, School of Cancer & Pharmaceutical Sciences, King’s College London, London, United Kingdom

    Contribution

    Conceptualization, Resources, Data curation, Supervision, Funding acquisition, Writing – review and editing

    For correspondence

    tony.ng@kcl.ac.uk

    Competing interests

    has received research funding from Astrazeneca and Daichii Sankyo. TN is a founder and shareholder in Nano Clinical Ltd, and PRB is a shareholder

    "This ORCID iD identifies the author of this article:" 0000-0003-3894-5619

• 1,277

  views

• 216

  downloads

• 2

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.