Key 🔐 Cancer Facts
Cancer is the second-leading cause of death worldwide with 10 million people dying 💀 from cancer every year.
More than 40% of cancer-related death could be preventable as they are linked to modifiable risk factors such as smoking 🚬, alcohol use 🍻, poor diet 🍟 and physical inactivity 🪑.
Almost at least one third of all deaths related to cancer could be prevented through routine screening, and early detection and treatment.
70% of cancer deaths occur in low-to-middle income countries.
The total annual economic cost of cancer is estimated at $1.16 trillion.
🖇️ Source: World Cancer Day
So far, 71 AI-associated (or AI-associable) devices have already received FDA approval with cancer radiology having the majority of approved devices (54.9%), followed by pathology (19.7%), radiation oncology (8.5%), gastroenterology (8.5%), clinical oncology (7.0%) and gynaecology (1.4%). The vast majority of these approved devices (>80%) regarded the complex area of cancer diagnostics.
The Pathology AI associated/FDA approved devices are:
ER APP Breast Cancer, PR APP Breast Cancer, Hot Spot APP, Invasive Tumour Detection APP and Metastasis Detection APP by Visiopharm A/S,
DeepDx-Prostate Connect by Deep Bio Inc.,
Paige Prostate and Paige Insight by Paige Inc.,
Galen Prostate by Ibex Medical Analytics Ltd,
Cobas® EZH2 Mutation Test by Roche Molecular System, Inc.,
Her2 dual ish dna probe cocktail,
Cintec plus cytology and VENTANA MMR RxDx Panel by Ventana Medical Systems, Inc., and
FoundationOne Liquid CDx by Foundation Medicine, Inc.
🖇️ Source: Artificial intelligence in oncology: current applications and future perspectives
After this brief introduction, today’s newsletter is going to be dedicated to AI Startups/Tools for diagnosis (digital pathology and imaging), screening, drug discovery and clinical trials in cancer research.
Cancer Diagnosis: Digital Pathology and Medical Imaging 🔬🩻
⚙️ Owkin is a French-American full-stack AI biotech that identifies new treatments, optimizes clinical trials and develops diagnostics using histology slides and omics. They are offering:
➡️ Multimodal patient data:
📎 Abstra will host a metadata catalog to enable researchers and data scientists to discover collaborators and datasets.
📎 Substra is a ready-to-use, open source federated learning (FL) software developed by Owkin—now hosted by the Linux Foundation for AI and Data—that enables the training and validation of ML models on distributed datasets.
FL—also known as collaborative learning—is a ML technique that trains an algorithm via multiple independent sessions each using its own dataset, in contrast to traditional centralized ML techniques where local datasets are merged into one training session, as well as to approaches that assume that local data samples are identically distributed.
📎 MOSAIC is a landmark research project to create a multimodal dataset with spatial transcriptomes from 7,000 patients in 7 cancer types. This is the largest spatial omics atlas to date.
Spatial-omics is an overarching term for different technologies that allows overlaying of -omics data onto tissue images. Beyond the identification of molecular subpopulations of cells, these approaches can provide information regarding the (spatial) localization of the identified subpopulations within the tissue of origin, their proximity with each other and with the extra-cellular matrix, blood vessels and other tissue components.
➡️ Subgroup discovery: They apply AI to multimodal, KOL-curated data to subtype patients and identify novel biomarkers to inform drug discovery, de-risk clinical trials and develop and deploy diagnostics in clinical practice.
➡️ AI drug discovery (for novel drug targets and drug positioning) and AI drug development (to increase the probability of success of clinical trials).
➡️ AI diagnostics: They pre-screen for biomarkers and predict outcomes—giving healthcare providers a fuller picture of a patient’s disease. For example,
📎 MSIntuit™ CRC is a CE-marked AI diagnostic that provides a prescreen approach for digital pathologists. MSIntuit is used for Microsatellite instability (MSI), a key genomic biomarker that plays an important role in the treatment of colorectal tumors (CRC) patients, from H&E/WSI (hematoxylin-eosin-safran-stained/whole slide images). On June 14, 2023, Owkin successfully validated MSIntuit™ CRC AI that is now integrated into clinical workflows via France's largest network of pathologists, the Medipath.
MSI is the condition of genetic hyper-mutability that results from impaired DNA mismatch repair (MMR).
📎 RlapsRisk™ BC is an AI diagnostic to help pathologists and oncologists determine the right treatment pathway for early breast cancer patients. RlapsRisk BC assesses the risk of distant relapse at 5 years of ER+/HER2- early invasive breast cancer patients, post surgery, from HES/WSI and clinical data.
Regarding WSI, The ACROBAT challenge aims to advance the development of whole-slide-image (WSI) registration algorithms that can align WSIs of breast cancer tissue sections that were stained with immunohistochemistry or haematoxylin and eosin.
On October 17, 2023, Servier—a global independent pharmaceutical group—and Owkin announced a partnership to use AI to advance and accelerate better-targeted therapies across multiple disease areas, including oncology. On August 10, 22, Owkin and ASCO’s CancerLinQ—a health technology company focused on improving cancer care quality and clinical research with real-world data and advanced analytics—announced a new research collaboration to use AI in order to analyze real-world oncology data with the aim to understand why some cases of metastatic non–small cell lung cancer are resistant to first-line immunotherapy. ASCOCancerLinQ has one of the largest and most-diverse real-world oncology databases that include de-identified data from more than 6 million patients with cancer, as well as de-identified data from a European research site. And in 2021 Sanofi inked a $270M cancer AI deal, to utilize Owkin’s AI-driven platform.
Owkin is founder-led by Thomas Clozel, MD, Oncologist, and Gilles Wainrib, PhD, Professor of ML, is trusted by 8 biopharmas, is first in class in AI diagnostics and well-funded with million dollars (a total of $304.1M) raised from leading biopharma companies (Sanofi & BMS) and venture funds (Fidelity, Google Ventures & BPI among others).
⚙️ Paige—a global leader 🥊 in end-to-end digital pathology solutions and clinical AI with the first Large Foundation Model using over one billion images from half a million pathology slides across multiple cancer types—is developing with Microsoft a new AI model that is orders-of-magnitude larger than any other image-based AI model existing today configured with billions of parameters. For this collaboration, Paige is incorporating up to four million digitized microscopy slides across multiple types of cancer from its unmatched petabyte-scale archive of clinical data, and will utilize Microsoft’s advanced supercomputing infrastructure to train the technology at scale and ultimately deploy it to hospitals and laboratories across the globe using Azure.
⚙️ Boston Massachusetts-based PathAI, provides AI-powered research tools and services for digitizing and analyzing pathology images in order to make safer and more affordable the sub-typing of diseases like breast cancer. For example, PathaAI has the pathology market’s first algorithm to use additive multiple instance learning (aMIL)—the AIM-HER2 Breast Cancer—that delivers automated digital HER2 scoring. MIL models enable spatial credit assignment such that the contribution of each region in the image can be exactly computed and visualized, to provide greater transparency for how AI predictions are made.
On November 30, 2023, PathAI announced the availability of ArtifactDetect on AISightTM2, PathAI’s digital pathology image management system. This product automates the detection of artifacts on digital pathology WSI and quantifies the extent of artifacts on any WSI. Artifacts are unintentional morphological features that do not have any histopathological relevance and can limit the utility of these images for review by pathologists, so they require replacement of the image which can create delays in slide review and case turnaround time. Moreover, PathAI and Roche announced a partnership in 2021 and the collaboration began with distributing PathAI’s research-use-only algorithms through the cloud version of Roche’s uPath software, with a focus on immuno-oncology for the treatment of multiple types of cancers. PathAI has raised a total funding of $255M.
⚙️ Deep Lens (US)—a digital healthcare company focused on a groundbreaking AI approach to faster recruitment of the best-suited cancer patients to clinical trials—has one of the world’s first digital pathology cloud platforms (VIPER, Deep Lens’ integrated cloud platform) that combines lab, EMR and genomic data to match cancer patients to clinical trials and precision therapies at the time of diagnosis, accelerating recruitment and bring game-changing therapies to market sooner. Paradigm—a clinical trials data and patient-matching platform—acquired Deep Lens.
⚙️ Proscia (US), that makes digital pathology software and AI applications for cancer diagnosis and considered a leader in digital and computational pathology solutions, has a new partnership with Datavant—a leader in helping organizations securely connect health data—that will provide life sciences companies with digitized pathology data to power the development of novel therapeutics and diagnostics. This partnership will bring together Datavant’s privacy-preserving connectivity technology and the largest US health data ecosystem with Proscia’s Concentriq® digital pathology platform. Proscia has raised a total of $72M.
⚙️ Merantix is a Germany-based AI research and incubator lab, building ML companies in various fields, and Vara one of its spinoffs has trained its AI model on more than a million mammograms—from partnering radiology offices and hospitals—to detect irregularities and signs of cancer with reliable accuracy. The company aims to develop a cloud-based, on-demand platform that will put its cancer-detection AI at the disposal of radiologists across the world. Vara raised a total of €14M.
⚙️ Aidence, an Amsterdam-based startup has developed a medical image analysis software based on DL, the Veye Chest, that is already deployed in 10 hospitals and helps radiologists detect and report pulmonary nodules on a CT chest image to detect lung cancer. Aidence and Heart&Lung Health (H&L)—a specialist in chest and cardiac radiology reporting—have partnered to provide the AI-powered scan for lung cancer screening in Tameside, Glossop and North Manchester Clinical Commission Groups (CCGs). Aidence has raised a total of $14M.
⚙️ Aiforia (Helsinki Finland, 2013) by analyzing images uploaded to its cloud is allowing researchers to detect any visible feature or pattern at scale—including in tissues and cells—in order to understand pathophysiology. Aiforia’s platform brings together AI and high-performance cloud computing and assists image-based diagnostics by providing efficient and scalable solutions, enabling new discoveries and clinical support with highly accurate and consistent data. On June 21, 2021, Epredia—a global precision cancer diagnostics company —announced that it has entered into a distribution agreement for Aiforia's AI based pathology software, in order to distribute Aiforia's portfolio of preclinical and clinical pathology tools globally. In 2022, Aiforia had its fourth CE-IVD marked clinical AI Model for Breast Cancer to its rapidly expanding portfolio of novel tools for cancer diagnostics (Prostate Cancer Gleason, Lung Cancer PD-L1, Breast Cancer PR, ER, Ki-67). Aiforia has raised a total of €21.2M.
⚙️ AI scans using MRI technology and AI by Prenuvo offer a revolutionary full-body scan with every system and nearly all organs, from head to toe, being analyzed using MRI and AI, that can detect health problems before they become severe. The $2,500 scan now available can detect more than 500 conditions.
Cancer Blood 🩸 Screening
⚙️ Freenome (US) is developing next-generation blood tests for early cancer detection—powered by their multiomics platform which combines deep expertise in molecular biology and ML—to identify cancer-associated patterns among billions of circulating biomarkers from tumor and non-tumor-derived sources. In particular, the Freenome platform uses ML to analyse genomic, transcriptomic, methylomic and proteomic data in routine blood tests to spot the earliest signs of colorectal and other cancers (i.e. pancreas), and is not limited to look for mutations instead it detects changes in gene expression, immune activity and cancer-related proteins among billions of circulating cell-free biomarkers.
Freenome has acquired Oncimmune, a UK-based immuno-diagnostics developer with a commercialised CE-IVD marked EarlyCDT Lung blood test and an autoantibody platform, and has already collaborations with Merck, Siemens Healthcare and Walgreens. Freenome, after closing a venture capital megaround (series D funding $300M from GV Roche), has now a total fundraising of more than $800M in just seven years.
⚙️ In 2020, GSK opened a £10M research hub in King’s Cross, London to leverage AI for the discovery of new drugs to treat cancer and other diseases. The aim is to partner with other companies using AI for the drug discovery process. During the multi-year collaboration, the team will use a 3D cellular model of a patient’s disease, to study (with imaging techniques) how tumor cells from patients undergoing treatment interact with immune cells. Then the data will be used in collaboration with the GSK AI team to distinguish between high- and low- risk patients and to look for ways to prevent resistance during treatment. This collaboration is based on a novel ML model that integrates multimodal data, genetic and molecular traits, tumor location, imaging and biomarker blood tests.
Cancer Drug Discovery and Development 🧪🧫⚗️
⚙️ GSK was the first pharmaceutical company in 2017 to participate in the "Accelerating Therapeutics for Opportunities in Medicine” (ATOM) Consortium—a public-private consortium that aims to cut preclinical cancer drug discovery from six years to just one by leveraging AI. For this project, GSK gave ATOM the chemical and the in vitro biological data for more than 2 million compounds it has screened. Today ATOM is a partnership between: Frederick National Laboratory for Cancer Research, GSK, Lawrence Livermore National Laboratory and University of California, San Francisco. The ATOM models and softwares are available on the ATOM Modeling PipeLine (AMPL), an open source, modular, extensible software pipeline for building and sharing models to advance in silico drug discovery.
⚙️ Atomwise is using DL to analyze molecules through simulation that eliminates time taken by researchers to synthesize and test compounds. One of Atomwise's top AI strategies is the collaboration with FutuRx to launch A2i Therapeutics to develop novel small molecule immune-oncology agents. A2i Therapeutics is a joint venture designed to utilize Atomwise’s AI platform (AtomNet® technology) to target ADAR1 (RNA-specific adenosine deaminase 1)—a key protein involved in controlling the innate immune response, neurological disorders, metabolic diseases and potentially a key target for immuno-oncology and neurological homeostasis. In 2022, Sanofi joined forces in a collaboration with Atomwise inking a $1.2 billion biobucks research collaboration.
Moreover, the same year the French drugmaker Sanofi SA partnered with UK-based AI-driven startup Exscientia to develop up to 15 drug candidates across oncology and immunology—in a deal worth up to $5.2 billion in milestone payments—and with Insilico Medicine (signing a strategic research worth up to $1.2 billion).
In 2020, Insilico—the AI company that pioneered the applications of the generative adversarial networks (GANs), reinforcement learning, transfer learning and meta-learning for generation of novel molecular structures for the diseases with known and unknown targets—also launched a preclinical research program focused on finding new treatments for brain cancer, and has brought on the former global program head of GSK’s computer-aided drug discovery unit to help run it.
⚙️ Nested Therapeutics a Cambridge-based VC-backed stealth biotech company that recently emerged from stealth with $125M to fight cancer, is using a multidisciplinary approach that leverages computational engineering, biology and a handful of niche fields in life science. On October 12, 2023, Nested presented the first preclinical data from its lead candidate, NST-628, a mechanistically novel non-degrading molecular glue that targets multiple nodes in the RAS/MAPK pathway. NST-628 is a potent, fully brain-penetrant, RAS MAPK pathway molecular glue inhibitor with efficacy in CNS tumor models.
⚙️ Numerate is another data driven drug design company applying AI to drug discovery. Numerate agreed to lead discovery programs for identifying clinical candidates in Takeda’s core therapeutic areas of oncology. Numerate was acquired by Valo Health—that uses human-centric data and ML anchored computation to transform the drug discovery and development process—on November 1, 2019.
⚙️ At Recursion Pharmaceuticals’ headquarters clusters of robots treat millions of cells per week with drugs, stain them with six dyes and then take pictures to capture and quantify as many morphological features as they can. By pushing these data through a ML pipeline, they hope to find relationships that are invisible to the human eye and to tease out clusters of effects that can guide their drug discovery. In 2017, Recursion (Utah US, 2013) announced the formation of a research collaboration agreement with Takeda, so that Recursion can utilize its discovery platform to provide pre-clinical candidates for Takeda's TAK-celerator™ development pipeline (TAKcelerator, is an in-house semi-autonomous accelerator founded by Tauhid Ali). In January 2019, Takeda and Recursion announced progress in their collaboration and in May 2020, Recursion entered into a licensing agreement with Takeda in order to develop TAK-733 in Hereditary Cancer Syndrome.
In particular, TAK-733 (REC-4881) is a clinical stage MEK (Mitogen-activated protein kinase kinase also known as MAP2K, MEK, MAPKK) inhibitor and is being developed to treat a hereditary cancer syndrome and related areas of oncology.
Recursion Pharmaceuticals has raised a total of $665M and has a post-money valuation in the range of $1B to $10B as of Sep 10, 2020.
🚨 So far, more than ten AI-designed drugs 💊 are already in or are entering clinical trials and one of them is by Recursion Pharmaceuticals: Undisclosed for HRD-negative ovarian cancer.
⚙️ Iambic Therapeutics (formerly known as Entos) in California announced in October 2023 the closing of an oversubscribed $100M Series B financing (for a total of $153M) co-led by Ascenta Capital and Abingworth, and also including new investors NVIDIA, Illumina Ventures, Gradiant Corporation, Shanda and independent board member Bill Rastetter.
Iambic has combined physics and AI to create a differentiated drug discovery platform that achieves a step-change in the speed and success rate for delivering best-in-class and first-in-class development candidates with:
➡️ NeuralPlexer: a protein-ligand structure prediction 3D physics-based equivariant generative diffusion tool,
➡️ OrbNet: an AI-accelerated quantum chemistry Graph neural network architecture based on quantum features,
➡️ PropANE: a multi-endpoint property prediction tool, and
➡️ Magnet: a generative molecular design tool.
Iambic has discovered so far two candidates to advance into the clinic:
📍 IAM-H1, a highly potent and irreversible tyrosine kinase inhibitor (TKI) that selectively targets HER2 and HER2 mutants, while sparing EGFR. And
📍 IAM-C1, a potential first-in-class selective dual cell-cycle kinase CDK2/4 inhibitor to address unmet needs in terms of therapeutic window and treatment resistance in cell-cycle-driven cancers.
⚙️ Lantern Pharma in US is a uniquely positioned life science company utilizing AI to analyze genetic signals and molecular markers for patient response to small molecules, thus allowing researchers to find clinical uses for validated cancer treatments whose development has been discontinued (drug repurposing). Regarding its pipeline:
➡️ On September 26, 2023, Lantern dosed its first patient in a phase I trial of its drug LP-184 to treat advanced solid tumors. A total of 35 patients will receive an LP-184 infusion on a 21-day cycle, for at least two cycles.
LP-184 is a small molecule with favorable CNS penetration, that utilizes the mechanism of action known as synthetic lethality, to exploit common vulnerabilities in solid tumor and CNS cancers with DNA damage repair (DDR) deficiencies. In addition, Lantern’s AI platform, RADR, has highlighted overlapping gene dependency profiles between glioblastoma tumorigenesis and sensitivity to LP-184, such as EGFR activation pathways.
➡️ On October 3, 2023, Lantern announced that the in vivo data highlighting the enhanced efficacy of its drug candidate LP-184 in glioblastoma were published in Clinical Cancer Research.
➡️ On November 30, 2023, Lantern announced that the FDA has granted LP-284 Orphan Drug Designation (ODD) for the treatment of high-grade B-cell lymphoma with MYC and BCL2 rearrangements.
Lantern Pharma went public in 2020, and had post-IPO fundraising, all totalling in $95M.
⚙️ In January 2023, AbSci became the first company 🥇🏆 “to create and validate de novo antibodies in silico” using generative AI. In April 2023, AbSci and M2GEN (now Aster Insights), an oncology-focused health informatics solutions company with the most advanced lifetime consented clinico-genomics data, announced a partnership to create new cancer medicines. In particular, Absci’s generative AI drug creation platform will tap into M2GEN’s clinical and molecular data set, ORIEN AVATAR, to accelerate the creation of onco-therapeutics.
🚨 Just this week, Absci announced a collaboration with AstraZeneca to deliver an AI-designed antibody against an oncology target. Absci has raised a total of $237.9M in funding over 10 rounds.
Cancer Clinical Trials 💊💉👨⚕️👩⚕️
⚙️ Turbine.AI has an AI solution to design personalised treatments for any cancer type or patient faster than any traditional healthcare service. Their technology models cell biology on the molecular level and can identify the best drug to target a specific tumour with; moreover can identify complex biomarkers and design combination therapies by performing millions of simulated experiments each day. Last month the company established a collaboration with Ono Pharmaceuticals to identify novel oncology targets and so far has raised a total of $34M.
⚙️ MENDEL.AI uses AI tools to analyse clinical data, including medical history and genetic analysis from cancer patients. Its goal is to facilitate clinical trials in oncology research by organising and analysing RWD, and matching patients to the right clinical trials. On October 11, 2023, Mendel.ai launched Hypercube, an AI-powered platform for structured and unstructured text data that allows a user to ask questions in free text and receive outcomes from specific patient data. Mendel.ai has $65M.
⚙️ Deep 6 AI, utilizes AI to mine medical records to accelerate finding and recruiting patients for clinical trials within minutes. By using Natural Language Processing (NLP) to read doctors’ notes, pathology reports, diagnoses, recommendations and to detect hard-to-find lifestyle data— such as smoking and activity history— Deep 6 AI’s software managed to find and validate 58 eligible patients in less than 10 minutes while a principal investigator — using traditional recruitment methods — found 23 eligible patients in six months for a biomarker for a non-small cell lung cancer trial.
On September 6, 2023, Deep 6 AI announced the launch of its genomics module to accelerate enrollment for precision medicine and oncology research. The software uses AI to mine genomics data across an entire electronic medical record (EMR) system and disparate genetic reports to find patients with specific genetic markers in real time. Deep 6 AI’s CEO, Wout Brusselaers, explained that out of 30 million patient health records from across the US, only approximately 100,000 records contained sequencing reports with extensive genomics data in a structured or semi-structured format, while approximately 500,000 records had rich genomics information buried in free-text clinician notes and reports. And the beauty of the Deep 6 AI genomics module is that it uses NLP to contextualize all genomic and phenotypical clinical data—structured or unstructured—to precisely match patients to trials in minutes rather than months! Deep 6 AI raised a total of $45.1M.
For more: Cancer in the time of artificial intelligence (2nd part).
Until next time 🍵☕🫖🍪,