Data limits
The most important news in healthcare: Tuesday, April 14
Introducing Morning Pulse. We deliver the biggest stories in medicine, pharma and biotech straight to your inbox.
White House Model Lowers U.S. COVID Death Projections
The pulse:
A model built by the Institute of Health, Metrics, and Evaluation (IHME) and used by the White House projects that the U.S. will have 69,000 COVID-19 deaths through August – down from an estimate of 82,000 on Tuesday and 178,000 just two weeks ago.
Why the change?
The shifts in projections are based on improved data availability, both within the U.S. and other countries. Model builders have an improved understanding of the relationship between deaths, hospitalizations, and the need for beds and ventilators.
Notably, the most recent shifts have not been affected by the recent implementation of social distancing: the model has long assumed that strict social distancing measures will be implemented through the end of May.
What does this revision mean for social distancing?
For now, nothing, according to Dr. Anthony Fauci, who said on Friday that “now is the time to actually put the foot on the accelerator.”
President Trump, in parallel, is rumored to internally be pushing for a reopening of the economy by early May. 60% of the public, according to a CNN poll, are uncomfortable resuming regular routines on May 1, but a political divide underlies those figures: 53% of Republicans are comfortable, compared to 23% of Democrats.
How reliable is a model that changes so often?
While the original model relied only data from Wuhan, the latest has incorporated domestic figures as well as reports from seven cities in Spain and Italy. Still, experts warn against over-indexing on any one projection.
The IHME model is viewed by some to be overly optimistic, given its reliance on Chinese projections and its assumptions on how effective social distancing will prove to be. As a result, several states and districts have built their own projections – officials in Illinois and New York, for example, base decision-making off a composite of four models. Their approach is supported by the IHME model creator, Christopher Murray:
What we’ve learned from 30 years of weather forecasting, even Netflix predictions for movies … you make better predictions when using multiple models.
Photo source: Los Angeles Times
Mixed early data for hyped coronavirus treatments
The pulse:
Data on remdesivir, one of the most hotly anticipated drugs for the treatment of coronavirus, was released in the New England Journal of Medicine. Of the 53 patients in the study, 68% improved.
Meanwhile, France’s drug safety agency reported that hydroxychloroquine, the anti-malarial drug frequently feted by Donald Trump, appears to have serious heart-related side effects when used to treat COVID-19.
Remind me what these drugs are?
Remdesivir is an antiviral produced by Gilead. It failed to have an impact on its initial target, the Ebola outbreak, in the mid-2010s, but was revived after showing promise in lab tests against the SARS-CoV-2 virus.
Hydroxychloroquine is an oral tablet that has long been used to treat malaria and various immune conditions. Its potential as an antiviral has prompted patients and physicians to ratchet up global demand and push for validated testing.
Let’s start with remdesivir — does it work?
Well, it drove clinical improvement in 68% of patients, but the study lacked a placebo-controlled arm, making it impossible to tell just how effective the drug actually is. Select medical professionals are cautiously optimistic, noting that deaths within the study are lower than would be expected in a set of untreated patients and that remdesivir’s safety profile seems strong. Others warn against drawing strong conclusions, given the limitations of the trial.
Gilead is expected to publish data on remdesivir use in both mild and severe Chinese patients in the coming weeks. Analysts expect the drug to be more effective in mild patients, given that they have less virus in their body for the drug to fight.
What about hydroxychloroquine — is it safe to use?
The most recent data does not rule hydroxychloroquine out as a treatment for coronavirus and does not comment definitively on its overall risk, but it does highlight the need for an analysis of the benefits of treatment vs. the probability of side effects (such as the noted cardiovascular issues).
Though President Trump has encouraged use of hydroxychloroquine and the FDA has granted emergency approval to allows hospitals to use it from the national stockpile, experts are encouraging physicians to reserve the drug for only their most severe patients.
What’s the bottom line?
Despite the flurry of data, remdesivir and hydroxychloroquine remain potential candidates for COVID-19 treatment — no more, and no less.
AstraZeneca unblinds lung cancer trial due to overwhelming efficacy
The pulse:
We’re taking a page out of John Krasinski’s playbook to bring you some good news – so good, in fact, that AstraZeneca had to tell us two years early. The pharmaceutical giant has decided to unblind a global clinical trial for its drug, Tagrisso, as it has shown overwhelming efficacy in the treatment of lung cancer.
What is Tagrisso?
Tagrisso is a drug currently approved for patients with a type of lung cancer that expresses a mutated version of a gene called EGFR. So far, it’s been used for patients with metastatic disease and for patients who have failed other therapies. Prior clinical trials have shown that for these patients, Tagrisso delays progression of lung cancer by about 7 months.
In the most recent trial of 682 patients, AstraZeneca was looking to see how effective Tagrisso would be as an adjuvant, or add-on, therapy for patients who have already had their tumors surgically removed. Notably, the study includes patients with earlier stages of lung cancer that have not yet spread outside of the lungs. This gives Tagrisso the potential to treat 60,000 more patients globally and expand its sales, which are already sitting at $3.19 billion from 2019.
What exactly does unblinding mean?
Normally in a randomized clinical trial, patients don’t know if they’re receiving the treatment being studied or a placebo – they’re “blinded.” Though this trial was supposed to last three years, AstraZeneca has said they’ll be letting patients know what arm of the study they were in after just 12 months. In this case, because the study results clearly indicate Tagrisso’s benefit, it would be considered unethical to allow some patients to continue receiving placebo.
Well that seems unusual.
It is, but José Baselga, the head of oncology research has called the results of this trial “unprecedented” and wants to share the results as soon as possible. It’s a bit too early to judge for ourselves, as trial data has yet to be released. AstraZeneca has promised to share the full results soon.
Rapidfire
Gilead has lost the latest battle in its ongoing infringement saga with rival Bristol Myers Squibb, expanding their total damages to $1.2 billion. Ouch.
A major shipment from India of 3.5 million tablets of the anti-malarial drug hydroxychloroquine arrived in the United States on Saturday. India is a major global producer of the drug, which is being used as an experimental treatment for coronavirus.
Public perception of the deeply unpopular pharmaceutical industry is improving, in large part due to optimism regarding coronavirus efforts.
Closing Thoughts
We’re inundated with COVID-19 news. Death tolls, accounts from hospitals, ventilator estimates, political assurances, economic data — the information is unyielding, unforgiving, and often discouraging.
Paradoxically, our ability to address the disease is limited by how little we actually know. How many people actually have the coronavirus? How do we make accurate models for a virus we’ve never seen before? Do we wait to test new treatments until we have enough patients to draw definitive conclusions, or do we try what we can with the patients we have now?
The goal of Morning Pulse is to relay healthcare news, coronavirus-related and otherwise, honestly — acknowledging the limits of what we know while contextualizing what we can learn. We hope to make medicine a little more accessible and a little less scary at a time when staying informed is more critical than ever.
Stay safe, stay healthy,
Mrinal and Serena