What is Personalized Medicine?

bullseyeA few years ago I was asked to teach a course to adults at the ASU Osher School of Lifelong Learning about the Emerging Era of Personalized Medicine. This was exciting because it would give me the opportunity to help empower these adults to better understand their health, the science behind what make them sick, and what scientists and doctors are doing to cure them.  This was also a challenging course to develop because only a few years ago personalized medicine wasn’t the common buzzword like it is today. In fact, in early 2014, the Personalized Medicine Coalition contracted a research survey that found that 6 in 10 of people surveyed hadn’t heard of the term “personalized medicine” (see all results of the survey here). Despite the public being unaware of this huge advance, in the past few years, scientists and doctors continue to evolve this concept and medicine isn’t just “personalized” but now it can also be described as “precision,” predictive,” “individualized,” “stratified,” “evidence-based,” “genomic” and much, much more.

So what is new about this type of medicine?  Of course since the days of Hippocrates, doctors have provided care to patients that take their “personalized” needs in mind. Based on the patient’s symptoms and their experiences, the doctor provides treatment. But what if two patients have the same symptoms but different underlying diseases?  A fever and a headache could be the flu or malaria. Or two people could have the same disease, like breast cancer, but the underlying genetic changes are different so that the cancer should be treated differently for each patient.

The current concept of personalized/precision medicine uses each person’s individual traits (genetic, proteomic, metabolomic, all the -omics) and harnesses our molecular understanding of disease for the prevention, diagnosis, and treatment of disease.

personalized-med2The ultimate goal of personalized medicine is to improve patient health and disease outcomes. The graph above shows how better understanding the genetic and molecular causes of disease can improve health at all phases of disease progression.

  1. Knowing the risk factors that cause of disease (either environmental, like smoking, or genetic, like the BRCA gene mutation) can help to prevent disease before it starts by eliminating the risk factors or providing additional screening to catch the disease early.
  2. Biomarkers that detect disease before major symptoms can be used to treat the disease early, which usually has a better outcome than treating a disease that has progressed further (think stage 1 versus stage 4 metastatic cancer).
  3. Once a disease has been diagnosed, the molecular understanding of the disease can help determine what treatment the patient should receive (see below for an example).
  4. Biomarkers can also be used to predict whether the disease will progress slowly or quickly or whether or not a selected treatment is working.

For all aspects of personalized medicine, there lies the promise to make an enormous impact both on public health but also on decreasing the cost of healthcare.

breast_cancerLet’s use breast cancer as an example of how personalized medicine plays out in real life, right now. For breast cancer detection, breast self-exams and mammograms are typically used.  With personalized medicine, we now have an understanding of one of the genetic risk factors of breast cancer – mutations in the BRCA genes.  Patients at higher risk for developing breast cancer because of these mutations can be monitored more closely or preventative action can be taken. In the past, breast cancer treatment focused on treated with non-specific chemotherapy and surgery. Although both of these treatments are still of value, now doctors also test for the presence of certain breast cancer genes like Her2.  If Her2 is present in breast cancer cell, the drug Herceptin that specifically targets this Her2 gene can be used to specifically kill those cancer cells. If Her2 isn’t present, this drug isn’t effective, causes negative side effects and wastes time and money when a more effective treatment could be used.  Once breast cancer is diagnosed, a patient would be interested in knowing how quickly their cancer will progress. This used to be primarily based on the stage of the cancer, where stage 4 cancers have spread to other locations in the body so the prognosis isn’t great. Based on molecular markers, scientists have now created panels of biomarkers (Oncotype DX and MammaPrint) that predict breast cancer recurrence after treatment.

These personalized medicine-based tests and drugs are incredible. However, this is a field that both holds considerable promise and requires lots of work to be done.  For every incredible targeted therapy developed, there are patients that are still waiting for the treatment for their disease or the genetic variant of their disease.  In future posts, I’ll talk a lot about both the promise and the pitfalls of personalized medicine.

If you want to learn more about personalized medicine, check out this YouTube video with a cartoon comparing treatment with and without the concept of personalized medicine.

What is risk? Absolute versus Relative

riskMy mom and I were talking this afternoon – we talk every day on my drive home from work (I celebrated the day I got Bluetooth in my car) – about Angelina Jolie.  It was difficult to miss the news this past week about her New York Times opinion piece describing why she decided to remove her ovaries and Fallopian tubes.  There have been a number of interesting articles both praising (here or here or here) or criticizing or clarifying her choice.  That’s not what I want to talk about and it’s not what my mom and I talked about.  What we talked about was risk.  Most stories talking about Angelina Jolie mention that because of the gene mutation she had, there was an 87% risk of her developing breast cancer.  Despite the fact that 87% is awfully specific (and based on limited data from a certain number of women with this mutation that were studied over time), what I want to focus on isn’t the number, but what the number refers to.  In particular, I want to point out that there are different ways of talking about risk – and this is important when reading about any scientific information in the news.

coin_flipLet’s start with a quick definition – risk is the chance that something will happen.  These are usually percentages.  There is a 50% chance when you flip a coin that it will land on heads.  The risk is 50%.  Of course, when applied to the chance of developing a disease, or having a particular treatment outcome, or surviving an accident, the numbers are a lot more difficult to calculate than a coin flip.  But they are also more confusing when describing the risk as well.

 

rosk_tableI’m sure you’ve read news stories that say something like “Drinking more than 3 caffeinated drinks a day increases your risk of a heart attack by 50%” (this is a completely fictional example!!!) Fifty percent. What a HUGE risk.  Except what they don’t tell you is that without drinking caffeinated drinks, your risk of having a heart attach is only 1%.  So a 50% increase means your risk only increases to 2%.  This is the difference between relative versus absolute risk.  50% is the risk relative to what the actual baseline risk, whereas the absolute risk tells you the actual chance of something happening.

Let’s look at another example.  “This new drug decreases the risk of blindness in diabetic patients by 50% over 5 years”.  This is promising news!  Except, again, the 50% is relative risk – what you want to know is what the chance of a diabetic patient going blind?  If the chance that a diabetic patient goes blind is 60%, then a decrease of 50% is huge. There is only a 30% chance of blindness now.  Ont he other hand, if like the previous example, the actual chance of going blind is 2%, the 50% decrease is less impressive.  This makes the decrease in risk no less important to the patients who take the drug and don’t go blind – but it does affect how you read a news story describing the effect of the drug and whether or not you may want to take an expensive drug.

Now let’s get back to Angelina Jolie. The actual risk for breast cancer in the general population over a lifetime is ~12%.  If you have the mutations in the genes (called BRCA1/BRCA2) that Angelina Jolie has, it increases the risk to 40-80%. This is the absolute increase in the chance of getting breast cancer.  And as you may notice – the risk has a range (based on a number of factors – family history, health history, etc that we’ll get into in another post).

So how can you be a more savvy reader? You can be tipped off to relative risk by phrases like “increased by”, “decreased by”, “more than” or “less than”.  This only tells us the difference compared to baseline, but gives NO indication of what that baseline risk is. Absolute risk, on the other hand, provides the best estimate of what the overall likelihood of something happening will be.

risk_cartoon