Thursday, September 29, 2016

Uncertain Health in an Insecure World – 94

“Precise”, or Good Enough?

During World War II, the term “good enough for government work” meant that something being produced in the war effort had passed the most rigorous quality standards. Government-issued contracts for war munitions, ships and aircraft required the highest level.

After all, G.I.’s lives were on the line.

Since the era of The Greatest Generation, this term has assumed a pejorative meaning, suggesting that the work involved is either poorly executed and/or overpriced. The Reagan era Pentagon defense buildup was plagued with reports of $400 hammers and $600 toilet seats on Orion anti-submarine aircraft, implying governmental tolerance for colossal waste.

After all, national security was on the line.

In 2011, the U.S. Center for Medicare & Medicaid Services (CMS) admitted to making $70B in improper healthcare payments. So far, the 2010-2020 Obamacare policy funding envelope has increased access to healthcare for ~20 million Americans at a projected cost of $788B – that’s $4M per year for each newly-insured person! Ironically, this law now exacts high cost accountability from healthcare providers.

After all, there are covered lives on the line. 
To be clear, the goal of Obamacare is not universal coverage. Nor is it better population health. Nor is it improved quality of life. Cynically stated, Obamacare is largely system change for change sake. And the nemesis of all such massive government initiatives anywhere around the world is anti-change bureaucratic entrenchment. In an election season such as this in the U.S. (unlike any other?), at the core of this “good enough for government work” policy reform is political advantage.

After all, presidential aspirations and legacies are on the line.

Approaching The End, the Obama administration has decided to not make expanded healthcare insurance The Main Thing. Instead, in January 2015, they moved to misdirection. In rolling out the Precision Medicine Initiative (PMI), the U.S. President (above) announced that, “Doctors have always recognized that every patient is unique, and doctors have always tried to tailor their treatments as best they can to individuals.” He added, “what if figuring out the right dose of medicine was a simple as taking our temperature?”

Compelling stuff!

However, even if you’re the U.S. President posing beside the DNA molecule, that is not exactly PM.

The Cancer Moonshot 2020, chaired by U.S. Vice-president Joe Biden (above), was also launched in January 2016. Since the 1960’s, space travel has been more about geo-political advantage than scientific reality. Apparently, now cancer scientists have The Right Stuff! But if/when funded, this particular cancer moonshot isn’t going to land on the surface of a cure by the end of the decade, no matter that as a public initiative, it’s uplifting… even inspiring.

It and the PMI are both largely popular misdirection. And as Shakespeare wrote, “there’s the rub.”

Brilliant scientists are (generally) not great doctors, and vice versa. Sure, science informs modern medicine. But medicine is not the hypothesis-driven art of The Possible.  There’s a reason that jurisdictions don’t license smart Ph.D.’s to also practice medicine.

That being said, Dr. Ricardo Sabatini (below) is one brilliant scientist… a physicist, to be precise.

If you listen to his April 2016 TED Talk, I’m sure that you’d agree. During his fascinating description of DNA’s complexity, he welcomed Craig Venter (another brilliant scientist… a pharmacologist, to be precise) onto the stage. But Dr. Venter wasn’t there in the flesh. Sabatini unveiled a 450 kilogram 262,000 page encyclopedia listing all the DNA base pairs in Craig Venter’s personal genome. As Dr. Sabatini read Venter’s code of life to the audience, he turned the page to chromosome 14 in volume 132… “ATT, CTT, GAT…” After the laughter died down, this serious scientist pointed out how missing just two letters in this particular genome position would cause incurable cystic fibrosis (CF).

Let’s all agree on something – reading a normal gene sequence aloud, even from the pages of such a marvelous book, won’t cure CF any more than reading IKEA assembly directions in Swedish will magically put the furniture together.

When the actual Dr. Venter (and his business partner Peter Diamandis) created Human Longevity in 2014, they dreamed of translating the genome into improved human health. They marketed this approach, and the code of life concept, as the logical PM path forward. Their company employed 40 brilliant data scientists and used machine learning to read, re-read, and then interpret The Book. This actually allowed them use genome information to predict the features of a human face! And on the face of it (Sorry!), the results (below) are pretty precise and quite amazing!

But exactly how does all this massive data and scientific brilliance get translated into clinical medicine?

In the New England Journal of Medicine (NEJM, August 25, 2016), Dr. David Hunter (M.D. and Ph.D.) provided his perspectives on ‘Uncertainty in the Era of Precision Medicine’. First, he pointed out that the National Research Council (NRC) definition of PM “refers to the tailoring of medical treatment to the individual characteristics of each patient.” He reminds us that the NRC views the word “precision” as colloquial, meant to imply accuracy and a modicum of certainty in an outcome.

Dr. Hunter boldly predicted that precisely the opposite will probably occur in clinical practice.

He noted that, “tailoring treatments may demand greater tolerance of uncertainty, and greater facility for calculating and interpreting probabilities than we have been used to as physicians and patients.” In practice, oncologists and cancer patients must contemplate the relationship between the clinical and genomic risks of outcomes (i.e., metastatic tumor spread, survival, etc.). These risks are not always symmetrical. Dissecting a NEJM study of 6,693 women with breast cancer, Hunter reported that 70+ separate gene-expression tests carried out by scientists had to be dichotomized into ‘high-risk’ and ‘low-risk’ to permit clinical trial data analysis. If complex continuous genomic variables cannot be understood by physicians, then genetic profiles cannot inform patient choices regarding toxic cancer chemotherapy.

Sometimes, high science must be dumbed down to be “good enough for clinical work.”
To actually save lives, omics testing information must be integrated with clinical prognostic variables and tumor immunohistochemistry results. Until ~15 years ago, these clinical-pathological correlates were considered “good enough for clinical work.” In most community-based medical oncology practices, the drugs of choice are still cytotoxic platinum chemo agents. Omics has helped scientists to rewrite The Book. However, knowing the code does not reveal the subtle plot twists, the protagonist’s weaknesses, or the tragic themes in the patient’s history. Simply reading the CGAT’s aloud from The Book’s pages leaves many clinicians in the audience bewildered.

Clinical experience, not big data, remains the domain of doctors.

Assessing medical value, and enacting care properly informed by the probabilities, requires a straight-forward presentation of the options (and tradeoffs) by doctors for patients to absorb.

Widespread backlash to government actions is often fueled by press coverage. You hear what you want to hear in a news story. So to be precise, we now know that the Reagan Administration gave >$10M in arms illegally to Nicaraguan Contra rebels in 1986 using the tainted profits of arms sales to Iran. But to be clear, the complex plastic shrouds on Lockheed P-3 Orion aircraft (below) flying long-range high-risk anti-submarine missions are not toilet seats.

The same eye-of-the-beholder rule also applies to PM. 

We in The Square know that good science requires Precision. We also know that Medicine is often messy.

And while potentially compatible, these two truths are not mutually exchangeable. 

No comments:

Post a Comment