Systems therapeutics represent a paradigm shift in drug development. Historically in medicine, drugs primarily addressed disease symptoms, even if a molecular mechanism action was well understood. However, the introduction of biologic treatments for multiple sclerosis and immune-mediated disorders in the late 1990’s led for the first time to development of drugs that more specifically interfered with the disease process. The increasing application of systems biology in drug development led to further insights into the mechanisms of diseases which in turn allowed for the discovery of additional novel drug targets to interfere with and slow or halt the process of disease. In addition, new insights into the mechanisms of disease led to further differentiation in diagnosis and treatments, with significant advances in targeting patients with specific pharmacogenomic profiles for stratification of drug treatment. At the same time, novel insights into the mechanisms of inter-individual variation in pharmacokinetics and pharmacodynamics have led to personalized drug treatments.

Systems therapeutics are interventions which are:
•  Personalized with respect to the selection of medicines and dosing
•  Directly or indirectly impact disease progression
•  Complex (incorporating patient status, comorbidities, etc.)

An understanding of systems biology has allowed greater insight into the processes that govern disease and has helped to usher in the era of precision medicine, where the right patient gets the right therapy at the right dose and benefits from the right response. Systems therapeutics go hand in hand with precision medicine in that they both include patient stratification for drug response and personalization of medicines for individual patients.

Predictive biomarkers are essential components in identifying those patients who can be stratified in regards to response and risk. Along with predictive biomarkers, pharmacodynamic/pharmacokinetic profiles of individuals are used to understand a patient’s drug metabolism profile, anticipate drug–drug interactions, and will be necessary to track response to treatment. As a result, it will be necessary to work with rational combinations of drugs and diagnostics to induce a change in disease processes, disease progression, and ultimately cure. In an ideal future, each patient may receive a unique treatment plan designed for them.

Systems therapeutics can be broken into two broad categories: stratified and personalized. Both stratification and personalization are components of precision medicine . Stratified medicines provide differential treatments tailored to specific groups of patients with individuals within the group receiving identical treatment.

Stratified therapies for this report were selected for analysis based on the following criteria: the medicines came with a recommendation or requirement on their label for testing of a specific gene, protein, or hormone prior to use in either the United States, Canada, Japan or Europe. Products where testing of viral genotype was required prior to use were also included within the analysis. The 80 unique, stratified therapies in the analysis span over eight broad therapy areas including oncology, central nervous system (CNS), antivirals, genetic disorders, inborn errors of metabolism, respiratory, immunology, and rheumatology.
Personalized medicine takes stratified medicine one step further as differential treatments tailored to individual patients based on their specific genome as well as their status (e.g., pediatric, elderly, gender). The goal of personalized medicine is to treat the underlying cause of the disease in the individual patient (e.g., by altering the genetic code of the individual), but, in addition, would include the use of pharmacogenetic screening to aid in predicting clinical outcomes for diseases with associated comorbidities, pharmacogenetic screening for drug–drug interactions in patients receiving multiple drugs, and identification of a patient’s unique pharmacokinetic profile to aid in drug selection and optimal dosing.

Benefits of precision medicine
Precision medicines – medical treatment that is personalized to the characteristics of each patient by stratifying individuals into subpopulations that differ in their susceptibility to a particular disease or their response to a specific treatment – have the potential to provide significant value to patient outcomes. Prior to the advent of precision medicines, all patients with a diagnosed disorder would receive similar treatment, with some individuals experiencing benefit and others harm and added expense. The availability of precision medicine now allows for personalized diagnosis and treatment, with patients selected for treatment based on individual traits and medicines developed to specifically target the biological processes behind diseases.
Overall, precision medicine can add significant value for patients and health systems by offering improvements in response rate and tolerability, dosing guidelines, potential drug–drug interactions, and effective management of complex diseases. In addition to assessing treatment effect, pharmacogenomic testing can also identify those patients at greater risk for adverse events. Use and adoption of precision medicine rely on the availability of predictive biomarkers to identify patients who are susceptible to a particular drug effect, be it beneficial or harmful. Response to historical, first-line therapies across a number of therapeutic areas are currently suboptimal, and the ability to specifically target pathways in a disease has the capability to offer improved efficacy to certain individuals who harbor specific genotypes.

The use of pharmacogenomic data to aid in treatment decisions can also have a beneficial effect for other healthcare stakeholders. A study (n=110) conducted in a hospital-based home health agency investigated the effect of using pharmacogenetic profiling along with guidance from a clinical decision support tool (CDST) to aid in treatment decisions for patients receiving more than three therapies with the goal being to reduce the risk of dangerous drug–drug interactions. The results of the study indicate that hospital readmissions and emergency department visits were decreased by 52% and 42%, respectively, 60 days post-discharge in the pharmacogenetic tested group. The economic benefits from the use of precision medicine can therefore be consequential; the case study above lead to health care savings of $4,382 per patient over 60 days.

The use of prognostic and predictive biomarkers to facilitate disease prevention also has significant potential. Cancer prevention programs can use pharmacogenetic screening for early detection. Many neurological disorders are challenging to treat, and early identification remains critical for optimal patient care and development of potential disease modifying treatments. In the future, it may be possible to identify risk of Alzheimer’s disease with predictive biomarkers, and at present, biomarkers are increasingly used in clinical trials for Alzheimer’s disease to support proof of efficacy. Research is also ongoing for prognostic biomarkers to predict a risk for epilepsy and to improve multiple sclerosis diagnosis and disease progression. Predictive biomarkers also provide an opportunity to identify and stratify asthma patients, and in future, will help to predict patient response to medication.

Characteristics of stratified and personalized medicines
Advances in the study of systems biology have transformed healthcare by moving the treatment paradigm away from symptomatic therapies to targeted disease pathways in specific individuals. In addition, improvements in speed, access, and affordability of personalized genomic testing since the realization of the Human Genome Project have allowed for the development of methods to ensure drug response, precise dosing, and minimize adverse drug reactions in individual patients.
Predictive biomarkers allow the stratification of patient populations based on their predicted response to a therapy.

For example:
•  Diagnostic tests identifying EGFR and ALK mutations in non-small cell lung cancer, KRAS status in colorectal cancers, and BRAF mutations in melanoma can help stratify patient populations into those who will have a more effective response to treatment.
•  57% of patients with metastatic melanomas carry a mutation in the BRAF gene allowing targeted therapy with BRAF inhibitor such as vemurafenib (Zelboraf) or dabrafenib (Tafinlar); response rates for treatment with BRAF inhibitors in patients with specific BRAF mutations ranged from 48–59% in Phase II and Phase III clinical trials.
Other predictive biomarkers can stratify patients according to risk. For example:
•  The availability of diagnostic tests to identify genetic variants of the cytochrome P450 (CYP450) enzymes, enzymes crucial for drug metabolism are significant tools to identify optimal drug dosing, inform risk of adverse events, and predict drug–drug interactions.
•  Of the over 200 drugs on the FDA’s Pharmacogenomic Biomarkers in Drug Labeling list, 35% of the biomarkers listed are for variations of a cytochrome P450 enzyme, and of those, 61% are for Cytochrome P450 2D6 (CYP2D6). For the purpose of the analysis for this report, we included those medicines that required or recommended testing for the variant on the label; however, labeling for CYP2D6 can be informative and not require testing.

Characteristics of stratified therapies
As of 2016, over 230 therapies across the United States, Canada, Europe, and Japan include pharmacogenetic information on their labels and the FDA’s list of therapies with pharmacogenomic biomarkers in drug labeling totals over 200 medicines. Stratified medicines included in the analysis in this report came with a recommendation or requirement on their label or based on regulatory professional society testing of a specific gene, protein, or hormone prior to use in either the United States, Canada, Japan or Europe. Products where testing of viral genotype was required prior to use were also included within the analysis and these included the direct acting hepatitis C therapies (e.g., sofosbuvir, daclatasvir).

A total of 80 therapies have been approved with labeling that requires or recommends pharmacogenomic testing . Thirty of the stratified medicines in our analysis received a pharmacogenomic biomarker that directed patient stratification post-approval, and 40 have received regulatory approval since 2011. The increasing availability of therapies with pharmacogenetic information enables healthcare stakeholders to take actions to optimize the healthcare system to better meet the needs of patients.

Oncology contributed the greatest number of molecules and represented 58% of the total number of stratified therapies.
•  The increasing use of prognostic and predictive Biomarkers to enable improvements in disease outcome, effect of treatment, and reduction in risk of toxicity is a positive shift in the treatment of many cancers.
•  Targeted cancer therapies, including those therapies with predictive biomarkers as well as molecules that interrupt cell processes, such as angiogenesis (blood vessel growth), represented 82% of all novel oncology therapies launched from 2010 through 2014.

Sixteen percent of the stratified molecules consist of antiviral therapies from the HIV and hepatitis C drug classes.
•  Globally, hepatitis C is associated with six main genotypes, each with different approved therapies and response rates.
•  Direct acting hepatitis C medicines are a success story for stratified medicines and provide high cure rates for hepatitis C.

CNS therapies made up 9% of stratified medicines.
•  The bulk of these molecules belong to the antiepileptic therapy drug class and biomarker testing is recommended or required based on the potential for adverse drug reactions.

Genetic disorders, which include inborn errors of metabolism (e.g., hyperammonemia, Gaucher disease) and other genetic disorders (e.g., cystic fibrosis, Duchenne muscular dystrophy), make up 13% of the number of stratified therapies.
•  Patient populations of these therapies tends to be minimal.
•  For example, there are approximately 30,000 cystic fibrosis patients in the United States and approximately 25,000 patients in Europe, while the disorder is rare and under-reported in Asia.
•  Even fewer patients have Gaucher Disease; in the United States there are approximately 6,000 patients.

A singular characteristic of stratified medicines is that there are a significant number of older, small molecules along-side more recently launched therapies and biologics.
•  80% of stratified medicines are small molecules.
•  The greatest number of stratified therapies have been approved or received biomarker designation post-approval since 2006, yet 31% of stratified therapies were approved before 2005 .

Diagnostic testing most often stratifies therapies based on predictive effect with fewer therapies requiring diagnostic testing to inform dosing and risk of adverse events.
•  83% of stratified medicines predict efficacy; within this group there were four stratified medicines that were associated with both efficacy as well as dosing and/or adverse reaction.
•  18% of stratified therapies require testing due to the prospect of adverse event or dosing recommendations.

The regulatory characteristics of stratified medicines show that the majority of the therapies were approved as new molecular or biological entities and 73% received priority regulatory review.
•  45% of stratified therapies are accompanied by a black box warning.
•  More non-orphan drug applications were associated with a black box warning than orphan drug applications.
•  Stratified therapies tend to treat serious, degenerative diseases (e.g., cancers) where providers and patients are prepared to accept risks in order to slow or halt disease progression.
•  Patient outcomes will hinge, however, on safety as well as efficacy, and patient monitoring and support during therapy is a key component of realizing the value of these therapies.

Components of personalized medicines
Currently available personalized medicines are gene therapies that can alter genetic information in specific cells in individual patients to treat or prevent disease.

Existing personalized gene therapies can be segmented into oncology and non-oncology products. Examples within oncology include:
•  Talimogene laherparepvec (Imlygic, formerly OncoVex) a genetically engineered viral therapy for the local treatment of recurrent melanoma.
•  Recombinant adenovirus–p53 (Gendicine), a viral therapy for patients with a mutated p53 gene for the treatment of head and neck squamous cell carcinoma.

Gene therapies that treat non-oncology indications include:
•  Alipogene tiparvovec (Glybera), an adeno-associated virus (AAV)-based gene therapy for the treatment of lipoprotein lipase deficiency (LPLD).
•  Strimvelis, a stem cell gene therapy for the treatment of severe combined immunodeficiency due to adenosine deaminase deficiency (ADA-SCID).
•  Neovasculgen, a vascular endothelial growth factor (VEGF)-based gene therapy for the treatment of atherosclerotic peripheral arterial disease (PAD), including critical limb ischemia (CLI).

Though not a gene-therapy, sipuleucel-T (Provenge) is a cell based immunotherapy for advanced prostate cancer that modifies a patient’s personal immune cells with genetically engineered fusion proteins which results in an increased ability of the immune system to fight prostate cancer.

Trends in the use of stratified medicines
Total use of stratified medicines in the developed markets, measured in standard units, has remained relatively flat since 2006. This is primarily due to the higher representation of small molecule therapies within the group, many of which experienced generic competition within the study period, balanced against the launch of novel oncology and antiviral therapies with low volumes .
•  4.5 billion standard units were used by patients in the 10 major developed markets in 2016, out of 4 trillion units for global medicines.
•  Within stratified medicines, older small molecules, particularly those in CNS and breast cancer, account for over 75% of the volume.
•  The perception that precision medicines are primarily biologic, disease modifying therapies is balanced by the reality that there is also significant use of older, small molecule medicines.

Use of molecules approved within the past five years has grown through 2016 and has begun to level off (see Exhibit 6).
•  Among the molecules with the highest volume since their approval in 2011 are therapies from the oncology and antiviral therapy areas, in particular, hepatitis C and chronic lymphocytic leukemia (CLL).

Oncology
Use of stratified oncology therapies across the developed markets has been growing since 2006 with a year over year growth rate of 4% in 2016.
•  The top 10 oncology therapies make up 94% of the volume of stratified oncology therapies; all are small molecule therapies.
•  Tamoxifen, anastrozole, and letrozole contribute the greatest share of volume and all are approved for estrogen receptor (ER) positive breast cancer. Among the top ten molecules by volume, year over year growth rate of breast cancer was 4.2% in 2016.
•  The oral chronic myeloid leukemia (CML) tyrosine kinase inhibitors imatinib, ibrutinib and nilotinib also contribute a significant amount to the volume of stratified oncology therapies, although growth has declined since 2010. Among the top ten molecules by volume, year over year growth rate of leukemia therapies was only 0.5% in 2016.
•  48% of stratified oncology medicines received regulatory approval since since 2011, and the CAGR was 80% in the period from 2012-2016.
•  Non-small cell lung cancer (NSCLC) was the only other therapy area represented in the top 10 oncology molecules by volume due to use of erlotinib.

Antivirals
Antiviral therapies include those HIV therapies stratified by risk of adverse events and the direct acting hepatitis C therapies that are stratified by viral genotype.
•  Overall volume of stratified antiviral therapies has declined since 2006, due to the decline of the stratified HIV therapies as more patients move on to newer medicines, such as dolutegravir, which is not stratified
•  In 2016, direct acting hepatitis C therapies accounted for 32% of stratified antivirals.
•  Hepatitis C therapies saw significant volume growth at time of launch, but uptake has since leveled off and will no longer drive significant growth in this therapy area in developed countries.

CNS
Current therapies for CNS are stratified by risk of adverse events, in particular, for therapies where cytochrome P450 testing is recommended or required. As a whole, these molecules are older, genericized, and less tolerable therapies.
•  Total units have been declining since 2006; the year over year growth rate for 2016 was -2.3% .
•  Antiepileptic drugs (AEDs) make up 98% of stratified CNS therapies.
•  Sodium valproate (divalproex semisodium; Depakote) has the largest volume among the stratified CNS therapies. The medicine has been launched globally with multiple formulations and has approvals across mood disorders, migraine, and epilepsy. As such, the volume of sodium valproate has remained relatively stable since 2006.
•  The use of other CNS medications have declined in large part due to the launch and continued uptake of more tolerable therapies, that, in turn, do not have labeling requirements for pharmacogenomic testing. Though effective, the risk of adverse events and side effects lead these older, stratified medicines to be prescribed less often.

Health system spending on stratified medicines
Spending on stratified therapies in the 10 developed markets totaled over $62.8Bn in 2016.
•  Spending across the developed markets increased at a CAGR of 20.5% from 2011–2016, driven primarily from growth outside of the United States and EU5; however, growth slowed to 3% in 2016. This change in growth rate is largely driven by a reversal in hepatitis C-driven therapy growth seen during the 2013-2015 period.
•  In the United States, spending on stratified therapies was $38.2Bn in 2016. Growth in the United States over the 2011–2016 period was also due to historically high price increases for branded and generic products on an invoice price basis, in particular, the high price points of oncology and hepatitis C medicines. Spending on a net price basis – reflecting off-invoice discounts, rebates and other manufacturer price concessions – is estimated to be $30.5Bn.
•  Spending on stratified medicines in the EU5 reached $14.4Bn in 2016. EU5 countries experienced significant growth in the past five years, with a CAGR of 11.3% over the period of 2011 through 2016. However, growth declined from 2015–2016 by -8.1%. This decline was due in part to the end of hepatitis C therapy uptake, as well as policymaker responses to unexpectedly high new drug spending in 2014 and 2015.
•  Although overall spending was dwarfed by the United States and the EU5, sharp growth was seen in the rest of the developed countries under study (Japan, Canada, South Korea, and Australia) with a combined CAGR of 19.3% over the 2011–2016 period.
Oncology and antiviral therapies made up the top five stratified therapies by spending.
•  The top five medicines by spending in the developed markets included the oncology products imatinib, rituximab, trastuzumab and the hepatitis C therapies ledipasvir/sofosbuvir (Harvoni) and sofosbuvir (Solvadi).
•  Spending on breast and blood cancers remained relatively flat from 2011–2016. Most notable are the launch and uptake of ledipasvir/sofosbuvir and sofosbuvir, which offer essentially an effective cure for hepatitis C. By 2016, these two medicines accounted for
33% of total stratified medicines spending.
•  Spending on biologics has increased from $10Bn in 2011 to $14.9Bn in 2016. The greatest growth was seen in small molecules, which accounted for $48Bn in spending in 2016.
•  Overall, combined spending for stratified oncology products declined from a high of 77% of the total spending on stratified medicines in 2011 to 47% in 2016 due to the rapid uptake of hepatitis C medicines. Antiviral agents accounted for 44% of spending on stratified medicines in 2016.

Oncology
Spending on stratified oncology products in the developed markets grew at a CAGR of 9.1% from 2011–2016 with growth rate of 12.7% from 2015 to 2016 .
•  Cancer remains among the leading causes of morbidity and mortality globally with over 14 million new cases each year and 8 million resultant deaths each year, projected to rise by about 70% through 2022 to 22 million.
•  Stratified medicines have had a significant impact on the treatment of cancer as treatments options and regimens are being tailored to individual patients by targeting specific molecular pathways and harnessing a patient’s immune system.
•  Half of the top ten stratified oncology therapies are used to treat types of leukemia, four treat breast cancer and one treats lung cancer.
•  The greatest spending was on the biologic therapies rituximab and trastuzumab, however, small molecule therapies imatinib and palbociclib (Ibrance) also contributed significantly to spending.

Antivirals
Spending on stratified HIV and hepatitis C medicines totaled $27.6Bn in 2016 with a CAGR of 58% from 2011–2016.
•  Spending on stratified antiviral therapies spiked in 2014 following the launch of stratified hepatitis C products in the developed markets. Previously, growth of spending on stratified HIV medicines was relatively flat.
•  In 2016, the spending on the top five antiviral medicines was approximately 90% of the total spending on stratified anti-invectives in the developed world, due in large part to price. The cost of stratified hepatitis C products is on par with some oncology products.

CNS
Spending on stratified CNS therapies totaled just under $2Bn in 2016 with an 8.6% CAGR since 2011.
•  Overall, the launch of generic competitors to extended release valproate semisodium (divalproex) led to a sharp decline in spending since 2008.
•  The antiepileptic drug class accounts for the greatest spending within CNS while the antipsychotic class has remained relative static.
•  The other CNS class, which includes dextromethorphan/quinidine (Nuedexta) and tetrabenazine, has driven growth since dextromethorphan/quinidine’s launch 2011 in the United States for the treatment of pseudobulbar affect.

Novel therapies
The late-stage pipeline for novel medicines is robust. As of November 2016 there were 2,240 total active therapies in late-stage development and an average of 40-45 new active substances are forecasted to be launched per year through 2021. These medicines will address significant unmet needs across therapy areas which include cancer, autoimmune diseases, and diseases of the CNS. Although the new medicines will include conventional mechanisms of action, they will also include novel, precise targeting of biological processes within complex diseases.

Patient selection (or stratification) by a means of a predictive biomarker is the defining criterion for a precision medicine or systems therapeutics. Predictive biomarkers are related to therapeutic response in that they identify patients who are susceptible to a particular drug effect, be it a response in terms of efficacy or in terms of adverse events. Other types of biomarkers include prognostic, diagnostic, and pharmacodynamic.

Among clinical trials with start dates in 2016, 350 studies selected patients using predictive biomarkers out of a total of 957 studies that included biomarkers.

In 2016, 4,726 clinical trials were started. Of these, 49% included some form of clinical biomarker. Of the total trials that included biomarkers, an analysis was conducted on those trials with biomarkers that measured a therapeutic effect or toxic effect in patients, not healthy volunteers. In total, 957 trials were analyzed to determine which studies selected patients using predictive biomarkers.

The percentage of biomarker trials that selected patients using a predictive biomarker has increased from 18% in 2014 to 37% in 2016.
•  In predictive biomarker trials with a start date in 2016, the greatest percentage was in early stage development, with 36% and 37% of trials belonging to Phase 1 or Phase 2 development, respectively.
•  The greatest number of trials that include predictive biomarkers is within oncology, followed by infectious disease and endocrinology.
Overall, there were 164 unique molecules being investigated among the clinical trials that used predictive biomarkers to stratify patient population in 2016.
•  The number of emerging agents has been increasing since 2014.
•  The greatest number of emerging agents is in oncology (55%) followed by infectious disease (11%).
•  Endocrinology and CNS made up 6 and 7% of unique molecules, respectively.     

(Courtesy :”Upholding the Clinical Promise of Precision Medicine” by Quintiles IMS Institute)