Artificial Intelligence Disrupts Drug Testing for Oncology

Sophisticated computer analytics can be tuned to predict the results of laboratory testing and thus render drug development both faster and cheaper. In pharmaceutical research, the quest for new drugs involved above all a very large number of lab tests. If, for example, basic research has identified a protein that has a role in tumor growth, often tens of thousands of active ingredients need to be tested in hopes that one of them will alter the protein’s function so that tumor growth will stop. At the same time, lab testing reveals some potential harmful side effects on the body. In other words, in order to find viable active substances, researchers need to generate huge quantities of data and review them for certain patterns. This sounds exactly like a computational task – and indeed, pharmaceutical research is about to change fundamentally and undergo unprecedented acceleration with the emergence of AI.

Especially R&D targeting complex systemic diseases stands to benefit. Large-scale assays for drug discovery are expensive and often impractical, particularly for the increasingly important physiologically relevant model systems that require primary cells, organoids, whole organisms, or expensive or rare reagents, as a recent article, Repurposing High-Throughput Image Assay Enables Biological Activity Prediction for Drug Discovery, demonstrates in Cell Chemical Biology. Oncology is morphing increasingly into a subspecialty of bioinformatics.

Time, effort and expense of conventional screening methods in the example described above stems from the fact that researchers need to test the interactions of each molecule under consideration with the protein they seek to influence, which translates into tens or hundreds of thousands of rounds of testing.

Deep Learning has been a prominent tool in toxicity prediction for some time. One hypothesis of the cited study was therefore that the pathway to a viable drug could be shortened significantly by having AI conduct specific preliminary selection. Instead of running a complete set of tests for each molecule, a different source of information might be capable of higher levels of automation: analyzing image data, which computers are already pretty good at interpreting. This led to the idea of a scalable machine-learning-based method predicting compound activity from images. The study hypothesized that data from a single high-throughput imaging assay could be repurposed to predict the biological activity of compounds in other assays, even those targeting alternate pathways or biological processes.

In the process tested here, cell cultures treated with the active substance under review are imaged using high-throughput microscopy. Preparation of cell lines is inexpensive and could also be performed by robots in the future. Cells are then dyed and photographed under the microscope. This is not only vastly more expeditious but also costs just $10-15 per picture.

Pictures are not evaluated by humans but by AI: learning algorithms receive data feeds from databases of already known substances and of their effects at cellular level. With this knowledge in store, software now scours through all new cell images and proposes a handful of the most promising options based strictly upon morphological changes observed in the cell structure. Quantitative information extracted from a three-channel microscopy-based screen for glucocorticoid receptor translocation was able to predict assay-specific biological activity in two ongoing drug discovery projects.

AI is aware that, whenever cells present a certain picture, laboratory tests have also shown a positive result in the past. The human eye is not good at systematically detecting those visual patterns. Which specific criterion AI uses to determine the outcome – whether it is the form of cell nuclei, the positioning of the cells with respect to one another, or other indicative patterns – remains as inscrutable and unpredictable as human intuitive processes, but its efficiency may be measured ex post. The study identified 125 promising molecules from a total of 500,000 candidates for an oncological application: hundreds of assays predicted by one image screen annotating half a million compounds. Subsequent random lab tests confirmed AI’s results in almost all cases. Repurposing increased hit rates 50- to 250-fold over that of the initial project assays while increasing the diversity of chemical structure of the hits. Connections discovered in the process are in many cases very surprising to human researchers as image-based models boosted hit rate and diversity in two drug discovery projects. Even if this method does not produce biological knowledge per se, it nonetheless provides starting points with excellent odds for further research. The study concluded that data from high-content screens are a rich source of information that can be used to predict and replace customized biological assays. It is expected that drug development will accelerate substantially as a result of involving AI-based screening and testing methods such as this, using wide varieties of pattern recognition. The instant study represented just a proof of concept justifying further work on image-based learning for drug discovery. It will also become more accurately targeted because problematic side effects can be predicted earlier. Of course, clinical trials will still not become obsolete until simulation of entire human organisms and their considerable varieties becomes available, which is why it will still take some time until the first AI-based drugs will come to market.


Algorithmically Tailored Immunotherapy: The Amazon Method of Curing Cancer?

“Customers who bought this item also bought…” Well, in the case of therapies for life-threatening diseases, Amazon’s familiar trope does not sound too helpful since it will frequently, and quite predictably, list “all of the above,” especially before insurance benefits have run out. And yet, it is a highly valuable if nascent computational concept, even though a long road still lay ahead before it may start to limit wasteful and disheartening trial-and-error in clinical reality.

The first effective oncological immunotherapy was approved just a few years ago. Today, patients with advanced skin cancer receive this therapy where medication supports the immune system in its struggle against cancer. Data analysis of several thousand skin cancer patients who were originally predicted to have only a few months to live shows that in about one-fifth of them, even two to three years after immunotherapy, tumors show no signs of return. This turned into a medical sensation that was all the more impressive as these drugs showed results not just with skin cancers but also with other types of malignancies, and now immunotherapeutics are approved also for lung, bladder and kidney cancers, adding more variants almost by the day.

The downside is that immunotherapy is effective only in about 20 percent of patients with most cancers. There are varieties, such as colon cancer, that are largely resistant to immunotherapy. Again, bioinformatics holds considerable promise to change this picture, since colorectal cancer is one of the most common types, with 1.4 million new cases per year world-wide. 

To be able to predict which combination of immunotherapy and standard therapy would be effective in a given patient, researchers breed in the lab mini organs, so-called organoids, which have been surgically removed from cancerous tissue. These organ-like structures are small tumors in their own right with the patient’s own genetic footprint. They are therefore ideal for testing the effect of various medications on the immune system of a specific individual. Data from these extensive tests will then be fed into a computer model that should ultimately predict which combination therapy would be most effective for each patient. A first review of a comprehensive study of over 600 colorectal cancer patients in Cancer Genome Atlas showed that infiltrating antibody cells of a tumor are very different depending on its origin, patient’s genetic predisposition, and tumor environment.

The number of mutations also varies greatly from tumor to tumor – from several hundred to several thousand per patient. This heterogeneity of one and the same tumor could signal a turning point in the search of cancer immunotherapy that fits all patients. It appears that hardly two colon cancer patients have identical mutation profiles. Therefore, breeding individual organoids that react very differently to treatment is essential.

Manipulating organoids by genetic and pharmacological means yields a lot of knowledge about drug treatment of tumors that is required to make them respond to immunotherapy. Technology for breeding organoids exists only since a few years and can be combined with high-throughput screening technology, which allows accurate characterization of organoids and determination of molecular profiles.

In order to derive targeted information for immunotherapy of cancers from bioinformatic analyses, an additional step is required: creation of tailored mathematical simulation models to process data feeds. This permits meaningful simulation of a wide variety of combination therapies. The number of possible combinatory permutations is enormous: in recent years, at least 70 new cancer drugs were approved, in addition to conventional chemotherapy still in use, and at least six new immunotherapeutic drugs. Algorithmic models can, in principle and depending on their quality, distill individualized combination therapies for each patient, creating a form of targeted precision medicine based on detailed knowledge of the specific mutation profiles of each tumor.

Key to that approach is the supply of the greatest possible amount of data from different clinical trials. Individual patient profiles need to be combined with data of similar patient groups and information about the medication; after evaluation, specific therapy recommendations could be expected to result. This process of advanced combinatorics is ultimately similar to the way Amazon evaluates customer interest from past purchases of individuals and larger groups; in oncology, scientists are aiming at specific recommendations for customized cancer therapy. Human oncologists would be overwhelmed with distilling and interpreting such data floods with any degree of accuracy, but combinations of AI and BigData methods can be expected to show the pathway toward the most effective therapy – not only for each type of cancer but also for each individual’s situation. With immunotherapy cost currently around $150,000-240,000 per year (and rising rapidly), accurate prediction as to what therapy combinations will be effective for which patient is not only key to survival for the individual but also for an already heavily strained health care system. The choices involved in the process inescapably demand answers to that most inconvenient of political questions: how much is temporary human life worth?