How much suffering and damage is acceptable today? - The unbearably slow transition to #PrecisionMedicine

"Primum non nocere" - The unbearably slow transition to #PrecisionMedicine

If one follows the rather slow adoption of new approaches to clinical development and also therapeutic decisions, in particular for chronic debilitation diseases, one may feel the hippocratic oath has come out of fashion. Sounds exaggerated and overly emotional? Think again!

Your chance for a false diagnosis in your lifetime is 100%:

How do we do in clinical development?

A BMJ analysis by Mendel and colleagues calls for greater transparency on ethics committee decisions to improve trial design in order to better inform patients about inherent risk, in particular in placebo guided RTC. Mendel and colleagues (Problems with ethical approval and how to fix them: lessons from three trials in rheumatoid arthritis) quote an earlier analysis as follows

A recent study reported that over 10.000 people with rheumatoid arthritis have been randomized to control groups receiving ineffective treatment in trials of biological disease modifying anti-rheumatic drugs, risking “ irreversible deterioration in condition.”

Is this tolerable in light of rapid progress of analytics capabilities?

Apparently it is, since contemporary trial designs accept RTC including placebos and consequences that include irreversible damage. For example @AbbVie just reported that Upadacitinib Demonstrates Positive Rheumatoid Arthritis Results in SELECT-COMPARE Trial What @AbbVie does not explain! A significant amount of patients was experiencing joint damage in both, control & study groups; joint damage today is irreversible! Off note, this trial is designed to scale and continue over the next years! Pls note that this quote only exemplifies, what is currently pursued by most companies in this field, including studies with biosilimars! I.e. this is not an @AbbVie specific problem, but includes regulators as well!

These trials and many others ongoing or in planning mandate the following questions

1. In general, can we not do better already today?
2. Is it acceptable that treatment decisions in conditions formerly called "rheumatoid arthritis" are mainly based on what comes close to "trial & error"?
3. Why is there no comprehensive reference database matching responders with drugs that work best for them?
4. Is this, only because we do not even have comparable, reliable and valid molecular sub-classification of chronic inflammatory diseases; see also Lancet review on Rheumatoid Arthritis?
5. Ever more "new" drugs for rheumatic diseases, i.e. IL-6 inhibitors and biosimilars push to market following the same, from my end outdated development path.
6. So why do we not see #PrecisionMedicine or better Precision Trials for these conditions? NOT!

Although a most recent publications claims a disruptive step in stratification of patients to the right drugs, I remain skeptical if this biopsy driven approach will ever make it to the clinic Rheumatoid arthritis meets precision medicine; I doubt that a biopsy driven approach will scale, not talking about the fact that we see unfortunately most promising biomarker approaches that aim to predict a responsiveness to a given drug fail.

As outlined in an earlier post, a solution to identify responders to a given drug would be in sight, but related activities are scares. In fact, until we see value-based compensation it seems #PrecisionMedicine is not attractive to most global players in Pharma & Biotech. Furthermore clinical R&D processes are poorly using the advantages arising from scalable omics, real-world data and #machinelearing #deeplearing.

Must not all stakeholders including regulatory authorities like #FDA and #EMA as well as patient organizations question ethical standards of our outdated clinical trial designs? Why do we allowdrug development and clinical decision to be dominated by outdated concepts, in particular re chronic debilitation conditions, while the solutions are at hands? Yes, the convergence of technologies enables us to solve these seemingly insurmountable challenges!

We meanwhile have multi-omics technologies at hand that over time could serve not only to decipher complex biological pathway interactions, events related to the induction of a pathology, but also may enable detection of minimal changes indicative of a desired or also unwanted, possibly toxic biological response much earlier in the course of time. See Personal Omics Profiling Reveals Dynamic Molecular and Medical Phenotypes and a related review by Eric Topol: Individualized Medicine from Prewomb to Tomb. Excellent review by Mike Snyder 2018 Integrative omics for health and disease and related video on omics technologies: Omics Advancing Personalized Medicine from Space to Earth featuring Mike Snyder @Stanford

In particular in "rheumatoid arthritis" a new test marketed by Crescendo Biosciences, Vectra DA?, only uses 12 markers, which enable detection of putative drug responders earlier than with conventional imaging technologies. Thismay seem a baby step in the right direction, but if reproducible for various drugs also in development could help to avoid irreversible damage with results of hinting to lack of efficacy of a give drug much earlier than ultrasound or X′ray:

Such methods, even though not conclusive, appear mandatory to be included when designing new trials, since “irreversible damage” cannot be tolerated as an outcome for a patient that participates in a trial. In fact, non-responders must be switched to an alternative treatment instantly and as early as possible, whatever technology is available to secure patient's interest; i.e. avoid irreversible damage!

The road towards #PrecisionMedicine and Precision Trials:

How could we design trial better, decide which treatment is the right alternative in a clinical setting? In general MDs are still left with trial & error re marketed drugs as well as in planning trials, but again, measures are at hand to change this. Although hard to believe in time of Precision Medicine, no validated test exists to predict which patient is actually responding to what marketed drug at which dose and time. This conundrum is being discussed in more detail in an earlier post. Drug efficacy or personalized dosing and related monitoring are far in the air.

Off note, 3 out of the 10 most successful blockbusters, i.e. Humira, Enbrel and Remicade fall in this trial & error category, which will now be complemented by new biosimilars that also come again without any patient stratification strategy.

The result of trial & error medicine is that most patients even today miss the early disease stage “window of opportunity” to treat to full remission, while payers may bear costs for up to 80% non-responder patients being treated extensively without any clinical benefit until the next trial & error circle starts, when inefficacy will be "diagnosed" by irreversible damage at worst. The way payers may try to cut cost with a "recommendation" that seems to lack any scientific foundation is not the way to go, "A Push to Lower Drug Prices That Hit Insurers and Employers the Hardest", but rather concerted efforts of the industry to identify "their" patients, i.e. the best responders, while sharing all data that enable a comprehensive Clinical Decision Support System (CDSS), that would ideally be driven by technologies including #machinelearning or #deeplearning, as well as a Point of Care Dx to eventually enable Precision Medicine 24/7 including drug dose, nutrition and related lifestyle management.

Most likely, similar with cancer diagnostics we must revise how we classify disease phenotype, since a disease may not be what you think it is. This article corroborates the necessity to come-up with molecular phenotypes and abandon our clinical classification systems; Molecular stratification of autoimmune diseases based on epigenetic profiles marks one step in the right direction.

In essence the message is a simple as put forward by William Robinson @Stanford in “Decade in review—technology: Technological advances transforming rheumatology” 

“Given the heterogeneity of most rheumatic diseases, the diverse molecular pathways mediating their pathogenesis and the multifaceted roles that these pathways have in normal and pathological states, advances in treatment are likely to require approaches that inte-grate genomic, transcriptomic, proteomic, metabolomic and autoantibody profiles, such as the recently described integrative personal omics profile, or iPOP.”

In light of the analysis by Mendel and colleagues cited earlier, I personally feel we must not stop at informing patient’s adequately about their risk when engaging them in a trial, but actually academia and in particular the biopharmaceutical industry in concert with regulators @FDA @EMA must engage in building an organizational and technical framework today, which will allow new trial designs to be developed fast and adopted a.s.a.p. These measures must put the safety, efficacy and also the best interests of any patient in the center of trial designs and biopharmaceutical development in general. Academia is not of the hook, in fact, patient care must come before publications and career planning; see Why Most Clinical Research Is Not Useful.

Technically we have all the tools at hand, i.e. multi-omics, sensors, increased computing capabilities, machine/deep learning, but what is missing is a focused effort to build a platform that actually makes trials not only safer, more efficient and less cost intensive, but also comparable, reliable and valid. Of note, many tests under development may serve to answer one of the many questions, like identification of patients responsive to anti-TNF; but what if the test is negative? There are no comprehensive solutions, which allow best informed decision medicine available today or at least being published as a proof of concept. Generating such a disease management ecosystem is doable, but we need to start building molecular maps of health and disease wisely, not least in light of immense investment necessary to generate the former. This endeavor may require the development of new standard operation procedures and logistics for data harvesting and analytics to avoid errors from the past; see also #PrecisionMedicine or "garbage in - garbage out"? “

The path from big data to precision medicine is all but trivial, but doable if we engage in vertically integrated flagship project, that are ideally designed for global scale, like this has been recently fostered by Keith Yamamoto in Precision medicine: Beyond the inflection point

Finally, there is accumulated evidence and growing awareness that change is needed now: as highlighted in one of my favorite papers by Sven Van Poucke et. al. titled: Are Randomized Controlled Trials the (G)old Standard? From Clinical Intelligence to Prescriptive Analytics 

We at Innventis therefore propose an alternative approach, which considers intra-indivual changes in a state of aggregation (personal molecular profile) to quite robust, which will also enable the delineation of drug-induce molecular profiles (DMP) the ideally will be predictive of clinical efficacy; in general, longitudinal multi-omics analysis of multi-omics profiles seems today the most promising way to go!

It seems the general concept already finds interest at least at one leading diagnostics company, i.e. Roche seeking a data scientist to analyze multi-omics longitudinal data in a rheumatoid arthritis trial. See summary of our vision @InnVentis presented a #PMWC18 @PMWCIntl

Join the discussion @The International Congress on Precision Medicine Beyond Cancer (PMBC2018) Munich October 15/16 2018 #PBMC18

Further reading:

• Why and how AMAZON will disrupt and dominate health care in 2025 - Will AMAZON buy Roche?
• Fixing a broken R&D model in Precision Medicine
• PRECISION MEDICINE
• SYMBIOTIC INNOVATION: A paradigm shift in R&D
• Reinventing Biomedical/-pharmaceutical R&D
• The 128 US$ billion (US only) Arthritis Conundrum:
• Will Google disrupt Medicine, Health Care

Disclosure:

InnVentis is a multi-omics/machine learning company was developed alongside the “Symbiotic Innovation” paradigm with the support of deep innovation GmbH, which is gratefully acknowledged. InnVentis goal is to build a vertically integrated B2B and B2C solution for precise diagnostics and therapeutic decision making & disease monitoring in real-time (supported by artificial intelligence) for major chronic inflammatory diseases; i.e. enable Precision Medicine.

For further discussions please join the LinkedIn group:

PRECISION MEDICINE INSIGHT