< PreviousPAEDIATRIC THERAPEUTICS 20 Pharma Business International www.pbiforum.net C hildren are continuously underrepresented in research and commercial product development. Meanwhile funding for adult conditions is often prioritised due to the higher expected commercial benefit, and feasibility of adult research, with limited clinical trial infrastructure in place for children. While discovery and development breakthroughs over the past few decades have presented treatments for a plethora of preventable diseases, children are often excluded from progress with most new treatments developed for adult populations. In the paediatric space, developing new treatments and offering effective and safe medicines is challenging, especially for rare diseases. This is the result of numerous factors, with a larger breath of considerations needed to be taken into account when for considering drugs and treatments for paediatric patients. Scientific, methodological, operational, clinical, ethical and logistical issues have limited and dissuaded the testing and production of products for the child population. This has seen children routinely given medications “off label”, which may be based on limited understanding of safety, suitable dosage and effectiveness. If one looks to developed nations like Canada, for instance, up to 80% of all medications prescribed in paediatric hospitals are administered off label and US researchers at Rutgers University have found children are increasingly being prescribed drugs off label without evidence for safety or efficacy. It has been well documented that children do not respond to medications in the same way as adults, with anatomical, physiological and developmental differences effecting absorption, distribution, metabolism and excretion. Further some conditions are limited to children, present themselves differently in children, and are more frequent in children. One particular barrier, making the development of paediatric treatments complex, is that within paediatric therapeutics there are varied sub-populations to address, based on development stages for instance, presenting a vast size variation, which may see one disease, its potential treatment and response to treatment appear differently in each sub-population. Additionally with these sub-populations comes a smaller eligible Problems for paediatric therapeutics The development and testing of paediatric therapeutics have historically been limited, with a habit of simply translating results from adults to children. 20-23.qxp_Layout 1 02/12/2019 10:25 Page 1Pharma Business International 21 www.pbiforum.net PAEDIATRIC THERAPEUTICS © Shutterstock /stockcreations 20-23.qxp_Layout 1 02/12/2019 10:25 Page 2PAEDIATRIC THERAPEUTICS 22 Pharma Business International www.pbiforum.net patient pool for studies, resulting in higher infrastructure requirements, with trials often needing to take place across several sites and even countries, thus taking longer to complete. The breadth of sub-populations also sees a need for different forms of treatments, and novel delivery methods. For instance infants, children, and adolescents will need different oral dosage forms based on swallowing capabilities, dosage requirements, pharmacokinetic responses and the need for taste masking. Paediatric oral dosages are challenging as they need to be administered easily but hold the flexibility to satisfy the varying needs of children in different stages of growth. This means different formulations that are age-appropriate is essential: liquids for infants and toddlers, mini tablets for older toddlers and young children, chewable tablets for young children and adolescents or capsules for adolescents. Primarily, design requirements for paediatric oral products are based on age, body size and swallowing ability. While liquid forms are useful for patients with swallowing limitations, are more readily absorbed than solid counterparts and enable flexible dosing, they tend to require special storage conditions, have short expiration timeframes, are associated with holding an unpleasant taste and thought to be harder to administer. Solid dosage dorms meanwhile have highly established manufacturing technologies and long-term stability. Easier to make, it is also simpler to mask taste with a coating. Of course with young children, however, they are difficult to swallow. Patient-centric approaches are therefore recommended for paediatric treatments. To provide incentives and create requirements for paediatric drug development, several legislative measures have been implemented globally. The US, for instance, has the Best Pharmaceuticals for Children Act (BPCA) from 2002, offering an extra 6 months of patent exclusivity to companies voluntarily carrying out paediatric clinical studies, alongside the Paediatric Research Equity Act (PREA) from 2003, which obliges companies to assess the safety and efficacy of new drugs in paediatric patients, enabling the FDA to necessitate paediatric studies of products likely to be utilised for a significant number of paediatric patients or which would have major benefits for children over current treatments. The programs, which push companies to conduct paediatric clinical trials and expand labelling to feature paediatric indications, became permanent after US Congress, in 2012, pushed through the Food and Drug Administration Safety and Innovation Act, aiming to make sure paediatric evaluations are conducted sooner in the drug development process. These have been somewhat successful, increasing the number of drugs with paediatric labelling with, as of April this year, 785 new paediatric labelling changes made because of BPCA and PREA. In Europe, meanwhile, the Paediatric Regulation, which came into force in 2007, requires pharmaceutical companies developing drugs that could be of interest for children to create a paediatric investigational plan (PIP) when applying for a marketing authorisation for new medicines or when developing current medicines in some cases. Since the regulation came into effect clinical trials have increased in Europe, by 50% between 2007 and 2016, and more medicines have been approved for children, while more information for patients and prescribers has become available. Regulations such as these have promoted the creation and strengthening of trial sites and investor networks for paediatric research such as the Connect 4 Children initiative. It has been highlighted however that while this legislation has seen increases in medicines approved for children, minimal progress has been made in diseases isolated to children, showing that paediatric therapeutics remains somewhat secondary to adult studies and treatments. Other moves are being made however to advance paediatric drug developments, such as a number of collaborations, one of the most recent being between site management organisation Penta ID Innovation and contract research organisation Cromsource. The partnership will see Penta’s paediatric infectious disease expertise utilised with Cromsource’s capabilities in managing worldwide drug development to address challenges in offering new therapies to children with unmet needs and see availability of new products to children accelerated. Overall the partnership will enable the sharing of scientific, therapeutic, regulatory and operational expertise to create a highly scalable clinical development capability for paediatric research. Investments are also being made, such as Pfizer’s £5 million investment into its Discovery Park site in Kent in the UK to improve its manufacturing abilities for paediatric medicines and patient-centric design. The funding will cover novel manufacturing technology that will allow scientists to explore innovative methods of making medicines more palatable with flexible dosing for children. The world is in dire need of more rapid development of safe medicines for paediatric patients. To succeed, the way in which treatments act in different stages of physical development must be considered alongside what dosage is best for a particular age group and how acceptable it will be to a child and their caregiver. More needs to be done to prioritise development of paediatric therapeutics independently, not as a secondary beneficiary of adult treatment development. 20-23.qxp_Layout 1 02/12/2019 10:25 Page 3Pharma Business International 23 www.pbiforum.net PAEDIATRIC THERAPEUTICS © Shutterstock /Photographee.eu 20-23.qxp_Layout 1 02/12/2019 10:25 Page 4ONCOLOGY 24 Pharma Business International www.pbiforum.net Artificial intelligence (AI), alongside machine learning (ML) and deep learning (DL), has integrated itself into many facets of contemporary society and various industries as a technology that, through algorithms, can predict, advise and learn based on data. The technology is establishing a place within the pharmaceutical industry and becoming commonplace in oncology as a highly beneficial tool with the potential to improve the outcomes of patients with cancer and optimise the work of pathologists, meanwhile AI approaches are being utilised in clinical trial design, radiology and facilitating tailored treatments. ML, DL and AI have opened up opportunities for digital pathology and analytics to better comprehend the connections between genetic mutations, disease and treatments. With growing medical records and patient repositories there are now huge amounts of information available to assist in this. With the large number of patients diagnosed with cancer and the amount of data created in their treatment, there is a high level of interest in how AI can be applied to improve care as a technology that can process huge amounts of data in seconds and analyse this data to extract useful information and produce effective treatment plans and best outcomes. One area AI is proving useful for is precision medicine. Where previously a physician’s choice of treatment or drug has depended on limited information and has been subjective to their judgement, AI can process a wealth of complex information, analyse available data objectively, and identify a pattern which can determine patient response to a certain treatment to ultimately help choose the right drug the first time. Oncologists have also highlighted AI’s use in predicting patients most at risk of complications as well as bolstering diagnosis accuracy. AI and oncology Artificial intelligence is a hot topic in oncology, with its applications spanning from improving diagnosis to precision and personalised medicine. 26 Á 24-27.qxp_Layout 1 02/12/2019 10:26 Page 1Pharma Business International 25 www.pbiforum.net ONCOLOGY © Shutterstock /anttoniart 24-27.qxp_Layout 1 02/12/2019 10:26 Page 2ONCOLOGY 26 Pharma Business International www.pbiforum.net Meanwhile AI developed at the Cleveland Clinic has used ML to pull together medical scans and electronic health records to create personalised radiation therapy doses for patients. This personalised therapy can deplete negative side effects and decrease treatment failures. AI’s use also extends into clinical trials, connecting qualified patients to trails when diagnosed, speeding up trial enrolment and opening up access to therapeutic alternatives. An additional key application of AI is in improving pathology, automating routine tasks with pattern recognition and augmenting the pathologist’s role, expanding diagnoses with deeper insight into treatment options and outcomes. Proscia’s Concentriq software for instance uses AI’s deep learning to analyse digitised slides of tissues. A series of disease-specific AI modules are in creation such as one which classifies and screens skin cancer biopsies after or prior to a pathologist’s review to reduce errors and boost lab efficiency. AI is already being used effectively in diagnostics to analyse radiological scans to assist in cancer diagnosis and outcome prediction. AI has been found effective for streamlining cancer screening and detection and holds promise for detecting radiographic anatomic features of malignancies beyond what clinicians can achieve. A Google Health team has built an AI system in the past two years, trained on over 45,000 CT scans, which is able to outperform human radiologists in diagnosing lung cancer. Google’s algorithm discovered 5% more cancer cases and had 11% less false positives than a human control group. Meanwhile BioMind AI system has reached 83% accuracy in predicting brain hematoma expansion compared to 63% in a team of physicians. In a similar vein, researchers from Beth Israel Deaconess Medical Centre of Harvard Medical School found that the analysis of data through DL decreased breast cancer diagnosis error rates by 85%. Elsewhere, MIT’s Computer Science and Artificial Intelligence Lab has developed a DL-based prediction model able to forecast development of breast cancer from up to five years in advance. The model was trained on mammograms and patient follow up data to discover patterns not obvious or observable to humans. In a further exciting breakthrough, scientists have used AI to recognise patterns in breast cancer that human analysis has missed and have uncovered five new types of the disease, matched to different personalised treatments. The study, led by a team from The Institute of Cancer Research (ICR) in London, uncovered that two of these types would better to respond to immunotherapy, and one was more likely to relapse on tamoxifen. Now the researchers are developing tests for the types of breast cancer which will be utilised to select patients for specific drugs in clinical trials, aiming to make personalised therapy a standard aspect of treatment. The researchers had previously used AI to find five types of bowl cancer. Oncologists are evaluating the application in clinical trials. The overall objective is to apply the AI algorithm to different cancers and present information for each type on their sensitivity to treatment, combatting drug resistance and paths of evolution. Despite its potential, the use of AI in oncology is still facing roadblocks. Data scientists are being presented with unstructured electronic health records, which are harder to process, and multiple data sources all structured for different purposes, while routine databases often lack the quality needed to be used by AI algorithms. The way in which we collect and use data must therefore become smarter. In Europe, GDPR enforcement is also proving problematic in the development of AI algorithms. Authorities however have been actively backing the approval of AI technologies with for instance the FDA working currently on a regulatory framework for AI technologies holding health applications. In fact the FDA recently approved the first cloud-based deep leaning algorithm categorised under medical devices, which can thus be used in clinical routine. 24-27.qxp_Layout 1 02/12/2019 10:26 Page 3© Shutterstock /Xray Computer 24-27.qxp_Layout 1 02/12/2019 10:26 Page 4PARTNERSHIPS 28 Pharma Business International www.pbiforum.net The knowledge base among pharmaceutical companies is enormous, so collaborations that make the most of the skills and expertise of a number of companies can have a huge impact on the sector. There are certainly numerous examples of some of the sector’s major names launching extensive projects alongside their colleagues and peers. The ability to merge the specialist areas of two or more individual companies ensure they can punch well above their weight on a global scale. In many cases, collaboration allows R&D to be commercialised under the auspices of a major manufacturer. There are always partnerships and collaborations among the pharma industry – with new ones being announced every month. Vifor Pharma announced in November that they would be holding a commercial partnership with Janssen Pharmaceuticals Inc (part of the Janssen Pharmaceutical companies of Johnson & Johnson) to jointly commercialise its new diabetic kidney Partners in collaboration Pooling skills is valuable in any industry, but the pharmaceutical industry, especially, has seen some high-profile collaborations in recent years. 30 Á 28-31.qxp_Layout 1 02/12/2019 10:29 Page 1Pharma Business International 29 www.pbiforum.net PARTNERSHIPS © Shutterstock /ESB Professional 28-31.qxp_Layout 1 02/12/2019 10:29 Page 2Next >