Why is neuroscience so hard?

Unlike the significant strides cancer research has made over the last few years, the field of neuroscience is often seen to be lagging behind. Although the prevalence of neurodegenerative disease such as Alzheimer’s disease and other cognitive dementias are high and growing as the population ages, current treatments are severely inadequate. Attempts made by pharma in drug discovery for neurodegenerative diseases have often led to spectacular losses with a 99.6% failure rate in clinical trials for Alzheimer’s disease alone.

There are often cited reasons on the challenges of drug discovery for neurodegenerative diseases which include:

  • Poor animal models (mice do not get Alzheimer’s)
  • Blood-brain barrier preventing drug entry
  • Poor understanding of pathophysiological disease processes (People are still arguing about whether tau tangles or amyloid beta plaques cause Alzheimer’s)
  • Poor diagnosis methods (dementia scoring is based on interviews with patients which can be rather subjective and you cannot do this with mice)
  • Focus on receptor-ligand model (the typical drug discovery approach which may work for other diseases but you never know with the brain, many compensatory activities can occur… it is an important organ.)
  • Patient selection for clinical trials (due to poor diagnosis methods, patients selected for clinical trials may not really have Alzheimer’s, or may be too far along in the disease for a drug to take effect).

Recently however, Biogen has reported pretty impressive results in its latest Phase Ib clinical trial for aducanumab, a human monoclonal antibody that targets amyloid beta (both soluble and insoluble forms) in Alzheimer’s disease. The trial had 166 patients that were split into groups assigned placebo or the drug at different dosage levels. Amyloid beta plaque levels were monitored by positron emission topography and at 54 weeks post-treatment, levels of amyloid plaques were significantly reduced in the mid to high dose-level groups. More significantly, patients in the highest dose group achieved a 5-fold improvement in the Mini Mental State Examination and Clinical Dementia Rating scores compared to those on placebo, suggesting treatment slowed the cognitive decline seen in Alzheimer’s. However, some severe side effects such as brain swelling were reported that led to discontinuation of treatment in some patients. Why Biogen has succeeded while others have failed may lie in their more careful enrollment of clinical trial participants. They made sure to exclude patients with other forms of dementia often misdiagnosed as AD and included those only in early stages of the disease. Aducanumab now advances to Phase III trials and if these positive results keep up, it would be a massive breakthrough for Alzheimer’s disease treatment and a humongous paycheck for Biogen.

In an effort to advance neuroscience research, several large funding initiatives have been launched in many countries including the US, Europe, China, Japan, Australia and Israel. They all have ambitious goals of creating a computer-simulated human brain, or establishing the functional connectome (in other words, mapping which electrical signals between neurons leads to specific changes in behavior). The European Commission has dedicated 1.19 billion Euro to its Human Brain Project over a 10 year period. President Obama has granted $100 million USD to the US BRAIN (Brain Research through Advancing Innovative Neurotechnologies) initiative for the year 2014 with a projected spending of up to $3 billion over 10 years. The jury is still out though on how effective these programs will be. In contrast to the Human Genome Project where the goal was clear (sequence the full human genome) and processes established, these programs simply consists of collecting large amounts of data in the hope that a pattern emerges which may reveal some fundamental principles of brain function or structure. Already, there has been some criticism that these large programs are limiting the availability of funds for other smaller research groups.

But clearly, understanding the brain is not a one-man task and an integrated approach may be what is needed to create some real advances. I suppose only time will tell.

*the cool image above is a scanning electron micrograph of cultured mouse hippocampal neurons (tan) on a layer of glial cells (brown) forming synaptic boutons from Thomas Deerinck and Mark Ellisman, the National Center for Microscopy and Imaging Research, UCSD


I have a PhD, now what?

When I was still in graduate school, as part of the student council, we organised a career workshop that coincided with the school’s orientation for new incoming PhD students. We invited an academic professor, a safety manager, a medical school lecturer, a sales and marketing professional and a team leader from big pharma, to speak on a panel and share their experiences. Overall, I thought it went well. We received good feedback from the students and even the invited industry speakers seemed to enjoy themselves. So I was surprised to hear some weeks later that the academic professor did not quite take to it. He believed the career workshop was wrongly timed and should not have occurred during the orientation of new students. Apparently because they should be focusing purely on science, career prep can come in the third or fourth year perhaps.

Wow, I thought. Really? It is true that the basis of a PhD is about extending scientific knowledge, but surely one has to acknowledge a PhD is a huge investment of personal time and money in lost working years. Not giving any thought to what career options lie after a PhD before even starting one seems pretty reckless to me. Yet it is what many students do. Many are clueless about what opportunities and paths lie in the road post-PhD. Many enter a PhD program hoping to become academic professors, not knowing only 10% of all life science PhD students get assistant professorships. So three full years can be spent being fully absorbed in one’s scientific project. You might become really good at growing cells, performing western blots, doing PCR, but so does everyone else. And when it comes down to applying for jobs, your cv lies indistinguishable in a pile of 70 other cvs from PhD graduates fighting for that industry job so they can get out of the seemingly never-ending post-doc position.

So no, I do not think we should be waiting towards the end of our PhD to decide one’s career. I think every PhD student should go into doing a PhD with their eyes wide open to the risks and opportunities. Below are various fields/positions that a fresh PhD graduate or even a post-doc with a few years experience can undertake.

  • Field application scientist/Sales
  • Senior Scientist (in a pharma/biotech)
  • Entrepreneur
  • Medical scientific liaison
  • Scientific writer
  • Product development
  • Business development
  • Regulatory affairs
  • Consultant
  • Government/Public policy
  • Patent law/IP
  • Technology transfer
  • Venture capital
  • Non-profit organisations

Of course, it may not be so easy to penetrate any one field. But being aware of the options help. And spending some time self-reflecting on what sort of job would suit you, and talking to people in these fields, would definitely arm you with the knowledge of what skill sets to develop during and after your PhD! I did not list academia though of course, it is an option. So before doing that next experiment, perhaps spend some time reading about these jobs. It is interesting to know for example that Medical Scientific Liaisons earn very high salaries (if that’s how you roll), but they require good knowledge on clinical trial organisation. So it might help to choose that PhD project where one gains knowledge on how a clinical trial is run with good access to patients and doctors. Or if you are interested in patent law, having the experience of applying for a patent or interning at a patent office might be of more relevance to you. Or if you want to be a scientific writer, get that writing experience by contributing to your weekly school magazine or starting your own blog. Taking actions like these not only help reveal if you really like doing something, but also make you stand out from the crowd when it comes down to job applications. So be proactive and yes focus on your science, but do not lose sight of your career.

The importance of good design

Recently read a book by Don Norman called “The Design of Everyday Things”. A pretty cool book on product design. At the time I was reading this, I was encountering some epic frustrations in the lab and it also provided some key insights I found useful in my situation. You know that feeling when everybody else reports successful outcomes with their experiments but when you try it, it fails and you don’t know why?

According to the book, when something you are using fails to work, it is usually not your fault. I found this thought rather reassuring 🙂 Most of the time it is due to poor design of whatever machine/tool you are using. This in stark contrast to what most people think. People are often quick to place the blame on others or themselves when things go wrong. In reality, there is a whole field assigned to human-machine interaction that focuses on psychology and cognitive science. People are not machines. We are a creative, random mess of emotions, thoughts and sometimes inexplicable actions. Engineers or scientist typically do not understand this. They thrive on logic and revel in mechanical design, and when a machine fails to work, they blame the user for not reading the instruction manual properly. Nobody reads an instruction manual in full detail. We have the habit of skimming through all the information, picking out things we think are important, and creating a concept of how something works.

So a good design has to accommodate this habit. It has to communicate its use simply and efficiently with what the author calls “signifiers” – something like a push/pull sign on a door. And it has to be shaped for its purpose, a concept termed an “affordance” – like how a teapot spout affords for water to be poured out of it. Feedback is also key. How many times have we pressed a button repeatedly not knowing if the machine has registered this? A simple light that turns on once a button is pressed would serve as good feedback (though still you have those annoying people that continue to push lighted buttons repeatedly.. recall waiting for traffic lights in Singapore). Often designers have a conceptual map of how something should be used, and users have to recreate this conceptual map gained only by interacting with the product. As such, communication via the product is key.

However, good functional design is often not enough. Psychology and emotion play a huge role in how we interact with objects. Its what makes us big fans of Apple, go crazy about Hello Kitty, and choose acai berries over strawberries at every opportunity. The author classifies user experience into three levels – visceral, behavioral and reflective. Visceral responses are natural attractions or repulsions to objects based on our senses. The attraction to clean, sleek contours and bright colours is something Apple often makes use of to capture audiences. Behavioral responses are shaped by previous experience and are also subconscious – the reason we choose Eppendorf or Qiagen over other lesser known brands for example, because we associate them with positive experiences. Lastly, the reflective level is the conscious assessment of the user experience which is often a combination of visceral and behavioural responses.

The book also had a good segment on failures, classified into slips and mistakes. Slips are unintentional mishaps usually due to memory lapses or task interruptions while mistakes were less benign, being the result of planned actions based on incorrect assumptions.

So why was I failing? To be honest its hard for a scientist to blame his/her machines, a PCR machine is pretty robust and reliable. Though some PCR machines have funny interfaces that can prove pretty frustrating. And often there are silly things that one does that can have an impact on your results, running your gel too fast because one is impatient for results for example can lead to fuzzy faint bands. It turns out it was poor reagent quality that may be leading to the failures. In the end, it is about finding out what went wrong. The reagents did not come with a disclaimer saying “there may be errors associated with our use”. It took quite a few experiments to find out what was going on. So design of not only products, but experiments is key to ensuring positive experiences. And certainly, it never helps to blame yourself!

The CAR T phenomenon

Let’s face it, we are all going to get cancer one day. We are living much longer, people are not giving up smoking, there is the occasional nuclear power plant leak, the ozone layer is still disappearing, and a host of other factors ensure that we are probably being exposed to an ever-increasing variety of carcinogenic substances. It’s no wonder drug companies are pouring so much money into cancer therapy development. There is a huge market and it also helps that cancer drugs are easier to get past regulatory authorities. Cancer patients are not particularly nit-picky about the side-effects of a drug if there exists hope that it can get rid of that inevitable death sentence.

Cancer therapy has undergone much progress in recent years. Where in the past, doctors had only surgery, chemotherapy and radiation to rely on, now there is a wide range of cancer-subtype specific therapies to choose from, mostly in the form of biologics. These include monoclonal antibodies, cancer vaccines, cytokine treatment, cancer-killing viruses, gene-therapy and even bacteria!

Currently, there has been a growing interest among pharmas on chimeric antigen receptor T cell immunotherapy or CAR T for short. Merck recently signed a US$941 million deal with Intrexon for their CAR T technology. Pfizer paid $110 million (with a potential to go up to $2.8 billion) for Celletics CAR T tech, and Novartis is also heavily invested, working with scientists from Penn Medical school that were the first to develop the technology. Various other biotechs specializing in this technology – Juno Therapeutics, Kite Pharma and Bluebird Bio – have also managed to raise incredible amounts of money within a short time-frame. The CAR T phenomenon as I call it, where everyone cannot stop throwing money into it.

CAR T therapy basically involves growing T cells from the affected patient (personalized therapy ftw) in culture, getting them to express CARs which target the cancer cells (usually CD19 is targeted that is primarily expressed on malignant B cells) alongside several moieties that increase the activity of T cells, and infusing them back into the patient where they target and kill cancer cells. Why is it being touted as the next big thing in cancer therapy? Primarily because of the impressive clinical trial results. One of the first in-human trials performed at the US National Cancer Institute in 2010 saw 6 out of 7 patients gaining partial or complete remission, with significant and sustained depletion of B cells and regression of adenopathy. Subsequent trials by Memorial Sloan-Kettering Cancer Center and University of Pennsylvania have all reported remission of malignancies lasting up to several months. This guy has been cancer-free for four years. In a larger trial sponsored by Novartis at the Children’s Hospital of Philadelphia, 27 out of 30 patients treated showed complete remission, with 19 remaining in remission as of Oct 2014. Several more trials are on-going to date, but the striking positive data seen so far have left many thinking that this may be the future of cancer therapy.

As usual, there are obstacles to overcome. Patients typically show rather acute adverse effects such as fever, fatigue, and hypotension that is usually accompanied by an increase in cytokine levels. The recent Philadelphia trial utilized another treatment tocilizumab, an anti-IL-6 receptor mAb to reduce this spike in cytokine levels. As such, there is a growing trend of combination therapies in an effort to increase the efficacy or reduce toxicity of treatment. There is a heightened interest in particular with immune checkpoint inhibitors – targeting PD-1 and CTLA-4 for example – that are said to increase T cell activity towards killing cancer cells. CAR T therapy also works primarily with non-solid tumours though several groups are now exploring their usage in solid tumour cancers.

Interestingly, genetics may also play a significant role in cancer immunotherapy. Patients with malignant melanoma treated with immune checkpoint CTLA-4 inhibitor3 for example, showed responses that varied with mutational signatures determined by exome sequencing. A researcher was quoted saying “The more mutated the tumor’s genome is, the more likely it is that immunotherapy will work.” A possible basis for the dramatic change in fate of pretty far along cancer patients that underwent CAR T therapy? With the excessive amount of funding going into CAR T, I hope we fully understand the specifics on how it works in future!


1. James N. Kochenderfer & Steven A. Rosenberg. Treating B-cell cancer with T cells expressing anti-CD19 chimeric antigen receptors. Nature Reviews Clinical Oncology 10, 267-276 (May 2013) doi:10.1038/nrclinonc.2013.46

2. CAR T-Cell Therapy: Engineering Patients’ Immune Cells to Treat Their Cancers – by National Cancer Institute

3. Alexandra Snyder, Vladimir Makarov, Taha Merghoub, Jianda Yuan, Jesse M. Zaretsky, Alexis Desrichard, Logan A. Walsh, Michael A. Postow, Phillip Wong, Teresa S. Ho, Travis J. Hollmann, Cameron Bruggeman, Kasthuri Kannan, Yanyun Li, Ceyhan Elipenahli, Cailian Liu, Christopher T. Harbison, Lisu Wang, Antoni Ribas, Jedd D. Wolchok, Timothy A. Chan. Genetic Basis for Clinical Response to CTLA-4 Blockade in Melanoma. New England Journal of Medicine, 2014; 141121104951001 DOI: 10.1056/NEJMoa1406498