WORLD CLASS BASIC, TRANSLATIONAL AND CLINICAL RESEARCH

Small cell lung cancer

Small cell lung cancer (SCLC) is an aggressive, highly metastatic, and incurable neuroendocrine cancer. This disease forms an important focus for the Cancer Biomarker Centre, and to contribute clinically relevant lung cancer research, CBC established a longstanding partnership with Prof Fiona Blackhall, a Professor in Thoracic Oncology at The University of Manchester and Lung Disease Group Chair at the Christie NHS Foundation Trust.

Patient derived preclinical models reveal novel biology of SCLC

The Preclinical Pharmacology (PP) team continues to expand the diversity of our biobank of CDX (currently >60 models). Several CDX were derived from patients before treatment with immune checkpoint inhibitors and following disease progression, facilitating discovery of tumour intrinsic predictive biomarkers for response to this recently licensed therapeutic option. In 2020, the team reported a new SCLC subtype driven by ATOH1, a neuroendocrine transcription factor not previously known to play a role in SCLC. They discovered that ATOH1 mediates a distinct transcriptional program, and both genetic and pharmacological targeting of this program increased cell death, implicating ATOH as a potential therapeutic target. They also exploited our pre and post chemotherapy CDX to discover a novel mechanism of acquired chemoresistance (an almost universal clinical occurrence), involving soluble guanylate cyclase that drives a nitric oxide-dependent signalling cascade culminating in activation of protein kinase G (see Schenk et al. Nat Commun. 2021;12(1):6652.) another potentially tractable target.

SCLC vasculogenic mimicry, perfusable tumour derived vessels

The PP team has continued to study ‘plasticity’ of SCLC cells, in this case exemplified by their ability to mimic endothelial cell behaviours via vasculogenic mimicry (VM). Having shown that VM vessels frequently contained red blood cells, they utilised a fluorescent lectin that interacts with glycoproteins in the basement membrane in vivo which labelled inner walls of both endothelial (CD31 positive) and VM (CD31 negative) vessels inferring perfusion with functional connectivity between VM vessels and the endothelium (Figure 1).

The first CDX model of an extra-pulmonary neuroendocrine carcinoma

Extra-pulmonary neuroendocrine carcinomas (EP-NECs) have limited treatment options and are poorly understood biology. In collaboration with Prof Juan Valle and Dr Mairead McNamara from the Christie NHS Foundation Trust, the PP team has generated and characterised the first CDX model of EP-NEC, and after histopathology (Figure 2) and RNAseq analysis, identified it as a Merkel Cell Carcinoma (MCC) EP-NEC, which prompted a review of the donor patient’s original diagnosis. This CDX recapitulated the biology and chemotherapy response of the donor patient’s tumour and holds potential as an avatar to guide future treatment.

SCLC Immunology

The Cells and Proteins (CAP) team has developed methods for ex vivo co-culture of immune cells (peripheral blood lymphocytes or Natural Killer (NK) cells) and disaggregated SCLC CDX cells to enable investigation of anti-tumour immune responses and resistance to immunotherapy (Figure 3). Early data suggest differential susCBCtibility between CDX models, as well as between the neuroendocrine and non-neuroendocrine CDX subpopulations within models, to NK cell-mediated killing. To complement these ex-vivo studies, an IHC assay against the NK cell reCBCtor NKp46 is being optimised for detection of tumour infiltrating NK cells. Generation of new CDX models (four so far), with banking of matched PBMCs from donors receiving immunotherapy, is also underway to progress co-culture research and study adaptive immune biology.

Subtyping SCLC via DNA methylation profiling of ctDNA

The recent description of molecular subtypes of SCLC with preclinical evidence of subtype vulnerabilities affords a ‘horizon view’ of personalised medicines for patients with this recalcitrant cancer. Given the paucity of adequate tumour biopsies, a blood test to subtype each patient will support this goal. The Nucleic Acids Biomarkers and Bioinformatics & Biostatistics teams have combined their expertise to optimise our genome-wide circulating free (cf)DNA methylation assay (T7-MBD-seq), with a bespoke bioinformatics pipeline and R package for analysis. They applied this pipeline to profile DNA methylation in >1,300 samples, including 593 cfDNA samples, across several studies including SCLC CDX and cfDNA from patients with SCLC (in collaboration with Prof Charles Rudin at Memorial Sloane Kettering Cancer Center, New York). Together they have developed a sensitive tumour/normal prediction classifier for disease monitoring and utilised differences in methylation patterns between the predominant molecular subtypes, based on NE transcription factors ASCL1, NEUROD1 and a double negative, to derive a molecular subtype classifier. These blood tests are now being validated in clinical cohorts.

Cancer of Unknown Primary – taking the U out of CUP

The Nucleic Acids Biomarkers and Bioinformatics & Biostatistics teams have also studied DNA methylation to develop a tissue-of-origin classifier to support treatment decisions in cancer of unknown primary (CUP) study in collaboration with Dr Natalie Cook, Christie NHS Foundation Trust. CUP describes a metastatic cancer cohort, with unknown primary tumour, making selection of beneficial treatment challenging. Tissue-of-origin classifiers were trained on methylation array data from 8892 samples, across 33 cancer types. These classifiers are in the late stages of development (Figure 4).

Bioinformatics for cfDNA methylation assays

The BBS team used a Nextflow framework to create a robust and reproducible data processing workflow enabling capture of quality control metrics and metadata. Using a novel approach, BBS converted data from methylation arrays (an unsuitable approach for cfDNA) including published data on SCLC and data from a range of tumour generated within The Cancer Genome Atlas consortium (TCGA) to comply with NAB cfDNA sequencing data. Machine learning was applied to detect presence or absence of tumour-derived DNA in plasma (ctDNA), SCLC molecular subtypes and tissue-of-origin for CUP.

ctDNA liquid biopsy biomarkers to direct therapy decisions: CACTUS, DETECTION and DYNAMIC trials in Melanoma

Guided by basic and translational research from Prof Richard Marais’ group at CRUK Manchester Institute, the Nucleic Acids Biomarkers (NAB) team, supported by our Quality Assurance team, have developed our portfolio of GCP-compliant liquid biopsy trials including CACTUS, DETECTION and DYNAMIC trial in partnership with Prof Paul Lorigan and Dr Rebecca Lee (UoM/CFT). The CACTUS trial (CirculAting Tumour DNA gUided therapy Switch) uses a ddPCR ctDNA assay to measure mutated BRAF levels that instruct treatment switch from targeted to immunotherapy for advanced cutaneous melanoma: 37 patients have been screened and validated ctDNA data returned to clinic within 7 days. The DETECTION trial (Circulating tumour DNA guidEd Therapy for stage IIB/C BRAF or NRAS mutant- positive mElanoma after surgiCal resecTION) opened to recruitment in late 2021. DETECTION involves ddPCR ctDNA analysis to detect early relapse/micro-metastatic disease and select patients for targeted therapy using a panel covering 3 BRAF, 4 NRAS and 2 hTERT mutations. NAB delivers the first validated DETECTION assay in January 2022; sequential samples from up to 900 patients are expected over the next decade. NAB are also validating cfDNA ddPCR assays to monitor tumour activity and burden (TAB) levels for the upcoming DYNAMIC trial (Circulating tumour DNA guided Adaptive BRAF and MEK Inhibitor therapy), which uses BRAF V600 ddPCR assays to monitor TAB to inform adaptive BRAF-MEK inhibitor therapy in Stage III unresectable/IV cutaneous melanoma.

cfDNA profiling to support drug development

The Nucleic Acids Biomarkers team are also working with UCB and CARRICK therapeutics to identify cfDNA biomarkers for patient response in their phase I/II clinical trials of UCB6114 (a mAb optimised targeting the human gremlin-1 protein) and the CDK7 inhibitor samuraciclib, respectively.

Biomarkers to inform immunotherapy trials

Reflecting unmet clinical need, the Cells and Proteins team continue to expand the CBC immune biomarker ‘toolkit’. Through a partnership with ThermoFisher, T-cell reCBCtor sequencing assays have been optimised to reveal T-cell reCBCtor clonality, diversity and convergence in blood samples. An early study using these assays involves a pilot set of longitudinal samples from NSCLC patients receiving immunotherapy/chemotherapy combination with ongoing clonal analysis of patient profiles by the BBS team.

CRUK Manchester Institute basic science biomarker discovery to the clinic

Working with Dr Santiago Zelenay at the CRUK Manchester Institute, the Cells and Proteins team has developed an assay to assess his COX_IS gene signature in FFPE patient samples using the clinically compatible Nanostring platform to validate its prognostic and/or predictive clinical value. Emerging data shows good intra and inter-assay reproducibility for analysis of clinical samples. Our ACED funded PhD student, also working jointly with Zelenay’s group, has analysed COX-IS and other immune/inflammatory gene signatures in early stage lung cancer resections to assess if there are associations with risk of disease relapse. A complementary COX2 IHC assay shows correlation between COX2 gene (Figure 5) and protein expression and is included in a multiplex IHC panel of 24 markers being established in collaboration with Prof Lisa Coussens (Oregon Health Science University) to enable assessment of expression levels and spatial distribution of immune cells within the tumour microenvironment and ultimately relationships with patient outcomes.

Community based blood sampling

The Cells and Proteins team and digital Experimental Cancer Medicine Team (ECMT), in partnership with Fiona Thistlethwaite from the Christie, assessed the feasibility of detecting 15 cytokines by ELISA in as little as 30µl blood collected using a micro-sampler device and optimised workflows for sample processing and analysis. This approach is being applied in the NOTION trial to evaluate home-based blood sampling kits that allow assessment of cytokines as an ‘early warning’ system of adverse events and cytokine release syndrome in patients receiving immunotherapy and advanced T-cell therapies.

Bioinformatics and Biostatistics across the Cancer Biomarker Centre

The Bioinformatics and Biostatistics (BBS) team integrates bioinformatics and statistical methods for its many and varied projects including specialised methods for single cell analysis to classify CTCs. They advise on the statistical powering of experimental designs including for the NOTION trial (described above) and the VALTIVE1 trial, the objective of which is to qualify plasma Tie2 for clinical decision making around anti-angiogenic agents in ovarian cancer in partnership with Prof Gordon Jayson (The Christie/University of Manchester), and the Accelerometer trial as part of our CRUK UpSMART Accelerator Award, coordinated by digital ECMT. The TARGET trial, led by Cancer Biomarker Centre alumnus, Dr Matt Krebs (The Christie/University of Manchester), matching patients with a broad range of advanced cancers to early phase clinical trials is now complete. Within TARGET, the Nucleic Acids Biomarkers and BBS teams have analysed and reported somatic mutations and copy number alterations across a 641 cancer-associated gene panel in a single ctDNA assay. The latest BBS data processing workflow was used to rerun all the ctDNA analyses producing a rich, internally consistent retrospective dataset now available to researchers to test hypotheses.

Digital solutions to support treatment for cancer patients

The digital Experimental Cancer Medicine Team seeks to digitally empower patients and healthcare professionals to innovate and design new cancer care pathways. They provide next generation patient cancer care through comprehensive data-driven evidence, enabling transformation of clinical decision-making, evolving the role of the patient and improving patient outcomes. This is achieved by listening to patients and healthcare professionals, understanding their needs and working proactively with them to develop ethical algorithms (AI) to support patient care, building digital solutions and evaluating technologies under clinical trial conditions (technology clinical trials).

Examples of digital ECMT’s distinctive research include: (i) The IN-HOME study, assessing feasibility of Acute Kidney Injury (AKI) detection in the patient’s home; Part A demonstrated feasibility and Part B is now opened to recruitment, evaluates potential for earlier diagnosis of AKI/change in renal function in cancer patients with intensive home monitoring. (ii) The A-EYE study, developing new AI methods to detect adverse retinal abnormalities associated with cancer treatment, now open to recruitment at the Manchester Royal Eye Hospital with >240 patients recruited. (iii) The NOTION study (with the Cells and Proteins team), enabling dry blood spot technology to measure cytokine levels in the home for early detection of immune related toxicities. (iv) The eTARGET tool, a digital solution integrating clinical and genomic data, histopathology data and corresponding images in a single portal to support decision making by Molecular Tumour Boards and deployed in two clinical trials being led from the Christie (TARGET National and CUP-COMP). eTARGET incorporated a link to the digital ECMT Cancer Trial Matching tool supporting clinical decision making by matching a cancer patient’s tumour genetic profile to optimal clinical trials. (vi) dECMT Artificial Intelligence (AI) researchers, business analysts and software engineers continued to improve the CORONET tool (COVID-19 risk in Oncology Evaluation Tool) using new datasets alongside input from the BBS team and our clinical colleagues. The CORONET team were recognised as “trailblazers in COVID-19 Research Response” at the NIHR Clinical Research Network Greater Manchester’s Evening of Excellence in November 2021. (vii) With colleagues in Italy and Spain, digital ECMT continued to lead our CRUK Accelerator Award, UpSMART to enable SMART Experimental Cancer Medicine Trials. Six Digital Healthcare Products (DHPs) were made available to the network of 24 sites (Figure 6) and with six more DHPs prioritised for the coming year. (viii) digital ECMT joined a consortium led by Vall d’Hebron (Barcelona) to secure EU Horizon 2020 funding for CCE_DART (Building Data Rich clinical Trials) with partner sites across Europe.