When 34-year-old Loise Wambui was diagnosed with depression eight years ago, the magic she had expected from the very first anti-depressants she was put on never happened.
Instead of lifting her out of depression, they caused her to balloon to obesity, break out with acne and turned her into a sleepwalking zombie.
Although, the therapeutic effects of the drugs didn’t kick in within the eight to 12 weeks that the doctor had promised it would take, she gave them the benefit of the doubt for two years.
Every time she complained that the drugs weren’t working, the psychiatrist remained adamant that the medication she was on was the best, and many of his patients were on it without complaints.
Eventually, her incessant complaints convinced him to prescribe something different, but the new drugs made her feel even worse.
Two weeks in, she rushed back to the doctor, fearing that the new drugs, which had increased her suicidal thoughts, would lead her to self-harm and then to the grave without delay. The psychiatrist gave her an “I-told-you-so” lecture, before quickly putting her back on the initial ineffective medication.
It was only when she decided to seek a second opinion, a year later, that she understood that when it comes to depression, patients may not respond to the first or second treatment prescribed, even though those very drugs do wonders for some patients.
Within a month of being put on a different cocktail of antidepressant and antipsychotic drugs, she felt like a new woman; the depression lifted and she felt mentally healthy for the first time in years.
Why will two people suffering from the same illness, have different responses to the same treatment? And is it possible for a patient to know whether a particular medication will work for him or her without having to swallow a single pill?
The field of personalised medicine has been providing answers in recent years. It posits that when it comes to treatment, one size doesn’t fit all and one therapy might work for one person and not another.
Physicians suggest that a patient’s individual genetic details should and can be used to predict the treatment choice that will work best for that patient.
Nowhere has this been more evident than in studies on the treatment of cancer, which is where researchers in the field of personalised medicine have focused attention.
When a person is diagnosed with cancer, the prescribed treatment often depends on the type of tumour, its location and the stage of cancer (whether it was caught early or at an advanced stage).
However, patients with the exact same tumour, with similar characteristics, may respond differently to the same treatment regimen – some with positive outcomes, while others fare badly.
With personalised medicine, also known as individualised or precision medicine, doctors use a person’s genetic makeup and other characteristics of the patient and their disease, to predict the best or most effective treatment for them.
For instance, a study published in the journal Nature Genetics in March describes the application of tailor-made therapy in the management of acute myeloid leukemia (AML), an aggressive blood cancer that develops in bone marrow cells.
There are 11 types of AML, each with distinct genetic features. Young AML patients can either get a stem cell transplant or chemotherapy. Stem cell transplants have better outcomes than chemotherapy, but still, one in four patients die from complications of the transplant, and another one in four experience long-term side effects.
Weighing whether to get a transplant, which may offer better cure rates, against the risk of early death that may come with transplant complications can be a tough decision for patients, their families and their doctors.
To take the guesswork out of the decision, precision medicine can be applied, with the patient’s individual genetic details being used to choose a treatment and predict the outcome. To be able to use individual genetic details to make precise treatment decisions, researchers built a knowledge bank using data from 1,540 AML patients, who participated in clinical trials in Germany and Austria. They combined information on each patient’s genetic features, treatment schedules and treatment outcomes.
This information could now be used to provide personalised information on the best treatment options for new patients.
It could also be used to calculate the benefits and risks for individual patients, thereby helping make the best treatment choice for each patient based on their profile. The researchers estimated that using personalised medicine, one in three patients would get a different treatment regimen when compared with current practice of treating all patients with the same disease using similar principles.
“The knowledge bank approach makes far more detailed and accurate predictions about the likely future course of a patient with AML than what we can make in the clinic at the moment. Current guides use a simple set of rules based on only a few genetic findings. For any given patient, using the new tool we can compare the likely future outcomes under a transplant route versus standard chemotherapy. This means that we can make a treatment choice that is personally tailored to the unique features of that particular patient,” said senior author Dr Peter Campbell of the Wellcome Trust Sanger Institute.
However, precision medicine requires large-scale genetic knowledge banks of up to 10,000 patients to predict the best treatment options for individual patients.
“Building knowledge banks is not easy. To get accurate treatment predictions, you need data from thousands of patients and all tumour types,” said co-senior author Dr Hartmut Döhner from the University of Ulm in Germany.
Such knowledge banks would also need to be kept up to date by incorporating newly approved therapies.
Even without a knowledge bank serving as a reference point, researchers have explored the option of using organoids, which are mini-organs grown in the lab using the patient’s tissue.
Doctors can create mini-organs with cancers of the oesophagus, stomach, pancreas, bile duct or liver, for instance. These artificial organs can then be used to determine which treatment will work best for that particular patient, based on the responses to treatment options tested on the mini-organs.
“In future, organoids can be used to find the best treatment combination for every patient, thereby increasing response rates, and reducing the side effects unnecessary treatment may cause,” said lead author Merel Aberle of Maastricht University in The Netherlands, in a study published in the British Journal of Surgery in January.
It’s not just chemotherapy where personalised medicine comes into play. In research published in the Lancet Oncology in 2017, scientists came up with a radiosensitive index that could be used to predict if a patient’s tumour would respond to radiotherapy. Being able to tell how sensitive one’s cancer would be to radiotherapy would then help doctors select the optimum dose for each patient.
“Just as every person is different, every person’s cancer might be different and therefore require a different course of radiotherapy.
“Yet, clinicians have been unable to make the distinction required to personalise radiotherapy,” explained researcher Louis Harrison, the chairperson of the Radiation Oncology Department at Moffitt.
The researchers took it even further and used scans to observe the genetic and cellular characteristics of the tumours in question, which would have a bearing on the outcome of the treatment. This would be better than biopsy, as the whole tumour could be sampled non-invasively and repeatedly.
Worth noting is that using genes to predict how a particular person will respond to a particular treatment regimen would not have been possible without the breakthrough that was Human Genome Project, an international collaboration led by US researchers, which took 13 years to complete.
In 2003, the scientists mapped the entire human genome, paving the way for the use of genes to tailor-make treatment for patients, or what is called medical genomics.
Whole exome sequencing scans a patient’s DNA for mutations in all 22,000 human genes at once.
Before sequencing of the entire human genome, experts would guess where in the DNA a problem might be, and look at each gene at a time, but this was not only costly, but also inefficient.
Dr Alexander Parker, a researcher writing in the Mayo Proceedings in 2015, noted that the genomic sequencing technology has become faster and more affordable, so that “instead of guesing where in the DNA the answer lies, we can look at all of our DNA at the same time and find the answer.”
In Kenya, where cancer deaths have been on an upward trend, with 16,953 deaths in 2017, there are currently two experts who can interpret genome sequencing results to give precise prescriptions.
One of them, Prof Francis Ndemo, who runs the Medication Therapy Management Clinic in Nairobi, says that genome sequencing tests that can be used to make individualised treatment decisions are not available locally, but can be done in South Africa. However, the cost would be prohibitive for many, which is one of the concerns that has been raised by some researchers.
Moreover, some scientists have also argued that the approach should go beyond genetics and include all other factors within and outside a person’s body, which have an impact on health. This may include environmental factors such as water sources, diet, a person’s social support network, or even on which side of the income divide they lie.
Researchers argue that without including these additional determinants of human health, the genomic information used to make treatment decisions would be incomplete.
Nevertheless, Prof Ndemo looks forward to the day when, after a local doctor or clinical officer has diagnosed and prescribed medication, there will be a clinical pharmacist to assess if the medicines are suitable for that particular patient, and if not, recommend the necessary adjustments.
He points out that most of the doses that local doctors rely on were settled on after trials with Caucasians, whose genotypes are different from those of Africans, and hence the medication may not be as effective.
If pharmacogenomics were to be integrated into primary healthcare, health workers could predict efficacy (how well the drug works for that patient), toxicity (negative side effects) and dosage (correct amount of drug to administer to a particular patient).
Moreover, testing which drug works best for which patient, would improve health outcomes and save lives.
Kenya should ditch one-size-fits-all approach in prescribing medicines
If Prof Francis Ndemo could do one thing, it would be to fast-track the application of pharmacogenics in Kenya’s healthcare, especially to manage chronic illnesses like cancer.
“The one-size-fits-all approach that is currently practised in our health system in prescribing medicines is harming and killing many patients,” says the pharmacotherapist, noting that using DNA information to prescribe medication would go a long way in improving patient health outcomes.
Problems like low and high dosage, adverse side effects, unnecessary drug therapy, and inappropriate drugs would be resolved with the application of pharmacogenics. This would require doctors to work closely with clinical pharmacists to make prescriptions.
Pharmacogenics goes way back to the 1950s, when University of Toronto scientist Werner Kalow was studying why some drug reactions seemed hereditary. Later, scientists discovered CYP genes, which metabolise drugs, and formed the basis for pharmacogenetic tests.
“CYP genes convert certain medicines into their active form and make them either more or less toxic. They also break down the drugs at different speeds or in some cases not at all, and this varies from person to person,” says Prof Ndemo, who has specialised in pharmacokinetics, or how a drug moves in the body when swallowed or injected.
The genes vary according to races, with Africans, for instance, having more diverse liver enzymes than Asians or Caucasians. This is because African ancestors were exposed to a wider range of plants, and as the world’s oldest ancestral group, their genes have had more time to evolve.
A study by Jim Kennedy of the Canadian Centre for Addiction and Mental Health found that people of African descent often require higher drug doses, because they have a wider variety of enzymes, which break down and eliminate drugs faster. Asians and Caucasians require lower doses to control conditions such as high cholesterol or hypertension.
People from the same race also differ in how they respond to treatments. Many medicines are broken down by more than one gene, so two or three genes can generate a drug response. Then there are factors such as age, weight, other health conditions, other drugs being taken, smoking, alcohol consumption and general diet, which also play a role in drug response, causing different people to metabolise drugs differently.
Knowing the genes that control drug reactions has become a useful tool for doctors and pharmacists who want to prescribe the most effective treatment. With pharmacogenomics, medical practitioners can match drugs to a person’s unique genome and prevent adverse drug reactions, which can be fatal.
Carolyne Abraham, a science author, predicts that if pharmacogenomics were universally available, it would cut the death toll from adverse drug reactions by as much as a third.
In the US, where scientists have made great strides in the field of pharmacogenomics, the Food and Drug Administration mandated that pharmacogenomics information be included on the labels of 300 medications, including painkillers, antibiotics and high blood pressure drugs, which are widely used in Kenya.