A lot of the news, whether about the healthcare area or just news in general, has to do with artificial intelligence – AI (as if everyone one on Planet Earth doesn’t already know that artificial intelligence is AI). I’ll get to the role of AI in healthcare later on in this submission, but there are several non-AI topics that deserve your attention, some optimistic and some not so optimistic. Let’s start with an optimistic bit of news.
A new Lp(a) drug?
A couple of week ago at the annual meeting of the American College of Cardiology, researchers from Eli Lilly announced that a Lilly drug, lepodirisan, could reduce levels of a particle in the bloodstream whose presence is significantly linked with a 25% increased incidence of heart attacks and strokes. An estimated 64 million persons in the US have elevated levels of this particle, called lipoprotein a – Lp(a). In persons with very high levels of this particle – perhaps 10% of the population – the risk of significant cardiovascular events doubles.
There is not a great deal of information out there about this particle. My Principles of Internal Medicine states that the Lp(a) particle consists of an apolipoprotein (a) molecule bound by a sulfhydril link to the apolipoprotein B part of an LDL particle. That’s about the sum total of what this worthy tome says about Lp(a).
Here’s what the Cleveland Clinic says about Lp(a):
“Lipoprotein (a) or Lp(a) is one kind of LDL or low-density lipoprotein. LDLs (cholesterol or fats) can cause issues because they can make speed bumps of plaque that slow your blood’s ability to flow through your arteries. Enough of these speed bumps can give you a blood-stopping blockage.
To make things worse, lipoprotein (a) makes your blood clot more and makes it harder for blood clots to break down. Blood clots are also like speed bumps. And inflammation from lipoprotein (a) makes it more likely that those speed bumps of plaque will break down, attracting more blood clots.”
Although a usual blood test measures levels of LDL as well as HDL cholesterol, the Lp(a) level is not generally part of a blood test. Cholesterol-lowering drugs, such as the statins, do not affect Lp(a) levels, and up to now no drugs have been known to deal with elevated Lp(a) levels, which affect about 20% of the US population. It has been estimated that only about 0.3% of the US population has had an Lp(a) test, and only 3% of individuals with heart disease have been tested. Perhaps that low level of testing is mostly due to the absence, up to now, of any effective intervention. Neither diet nor exercise have any beneficial effects in lowering Lp(a) levels.
However, there appears to be a possible breakthrough. The drug mentioned above, lepodirisan, is thought to reduce levels of Lp(a) by 94% following a single injection. According to the Lilly researchers, the effects of the drug last about six months, and there are no significant side effects.
Eli Lilly is now conducting a clinical trial to assess whether lepodirisan can prevent or reduce heart attacks and strokes. Results are projected to be announced in 2029.
Dr Steven Nissen, a cardiologist at the Cleveland Clinic whom we have cited many times in these postings, strongly urges his patients to be tested. Because the presence of Lp(a) is determined by genes, patients need to be tested only once. Dr Nissen says to those of his patients who test positive, “You have a disorder that has serious implications. I want you to take every risk factor that you have off the table.”
The first non-opioid pain medication to get the FDA nod in 20 years
From 1999 to 2022, according to the CDC, nearly 727,000 people died in the US from an opioid overdose, including both prescription and illegal opioids. To address the opioid epidemic that is plaguing the US, it has become increasingly important to find effective non-opioid pain management strategies.
A new drug that meets those qualifications is Journavx (suzetrigine), from Vertex Pharmaceuticals, which got FDA approval on January 30 this year. Suzetrigine is the first non-opioid analgesic to win FDA approval in more than 20 years. The FDA has stated that it is prioritizing the development of non-opioid pain treatment as a way to make pain treatment available without exposing patients to the risk of addiction.
Journavx / suzetrigine was evaluated in two randomized double-blind clinical trials with 874 subjects who had recently experienced surgical procedures. One of the drug trials followed abdominoplasty surgeries – known colloquially as a “tummy tuck” – while the other followed bunionectomies, an operation in which bunions are surgically removed. Clinical trials of analgesics are intrinsically difficult, since the placebo group is subjected to procedures that are known to result in acute pain, for which they receive no pain relief. However, all participants were able to use ibuprofen if the pain became too great.
“Both trials demonstrated a statistically significant superior reduction in pain with Journavx compared to placebo,” the FDA said in its press release. Note, inasmuch as the placebo group could use ibuprofen, Journavx could be said to demonstrate significant superior reduction in pain compared to ibuprofen as well.
Journavx / suzetrigine, however, is not a cure-all. It is meant for moderate-to-severe acute pain – pain that starts suddenly, often from trauma or surgery, and is expected to last less than three months. This means, based on the current evidence, that it would likely be used primarily in the hospital setting and only for a few days. That said, the drug will be available by prescription as well. And although the medication is not intended for chronic pain, the clinical trials did show that it had some efficacy in treating diabetic peripheral neuropathy (nerve damage that can cause numbness in extremities) when compared to a standard treatment.
Journavx / suzetrigine works as a sodium channel blocker. When we get an acute injury, pain-sensing nerve cells in the injured area respond by sending nerve impulses as a signal to the brain. Those signals are highly important. They give us valuable, sometimes life-saving information about the injury – where the injury occurred, and how bad it is.
Those nerve impulses are produced by molecules called sodium channels, which are bound in the membranes of nerve cells. The sodium channels act like gates, and when they open, they allow charged particles called sodium ions to enter the nerve cell. The electrical currents of the sodium channels create the nerve impulses that carry the pain signal to the brain. Sodium channel blockers are substances that stop sodium channels from operating, preventing the nerves from sending pain messages to the brain as effectively, which can reduce or eliminate pain. Journavx acts within the peripheral nervous system to block a specific sodium channel labeled Nav1.8, which is responsible for sending pain signals to the brain. The medication works by reducing the pain signals before they can reach the brain.
In addition to being nonaddictive, suzetrigine does not cause nausea or drowsiness, which are common issues with opioid medications. The studies found that the most common side effects of suzetrigine were itching, muscle spasms, and rash.
People should not take Journavx if they take certain medications, such as erythromycin, an antibiotic, and verapamil, a blood pressure/angina medication. Food or drink that contains grapefruit can also interfere with Journavx and should be avoided. Journavx can also increase levels of creatine phosphokinase, a blood enzyme which can be a measure of tissue injury in the muscle, heart, and brain, perhaps resulting in a false positive diagnosis.
According to Vertex, Journavx may temporarily reduce the chance of becoming pregnant, but women who use contraceptives should continue using them while being treated with the medication.
A brief interjection
What strikes me as particularly interesting about the lepodirisan and Journavx development is that in both cases the research was specifically focused at the molecular level – in lepodirisan, on one specific particle in the bloodstream, the one called lipoprotein (a), and in the case of Journavx, on a specific sodium channel whose function is to send pain signals to the brain. This narrowly-focused research is markedly different from the traditional pathway of drug development, which has been based on more general observation of the drug’s effects. For example, it was observed quite a long time ago that a tincture of willow bark relieved aches and pains. From that observation, it was discovered that salycilates (such as aspirin) were effective analgesics. (Salycilates got their name from salix, the Latin name for the willow tree.) Similarly, antibiotics were developed from naturally occurring molds. What these drugs did on a cellular and molecular level was discovered later on, but the effects were observed prior to any understanding of how the drugs worked.
Drug development has entered a new era. Researchers identify the specific target at the molecular and cellular level, and then find an agent that can specifically address that target and bring about the desired effect. I find that highly promising.
Some progress that may perhaps lead to treatment for atrial fibrillation
Atrial fibrillation, frequently referred to as atrial fib, affects more than 5 million Americans, but as Dr Patrick Ellinor, a member of the Broad Institute of MIT and Harvard and Mass General Brigham and professor of medicine at Harvard Medical School recently said, “Atrial fibrillation is an incredibly common disease, yet we have very limited pharmacologic therapies because we still have a primitive understanding of the molecular mechanisms involved.”
My attempt to shed some light on the research that Dr Ellinor and his colleagues are working on gets into the weeds a good bit. I apologize in advance and give you license to skip over the complications.
“Fibrillation” refers to twitching of individual muscle fibers acting without coordination. Atrial fib is the terms used to describe this kind of random and purposeless muscle twitching in the atria of the heart, the atria being the smaller heart chambers that take in blood from the arterial system and pump the blood into the ventricles. Ventricular fibrillation is a far more serious condition, which can lead to death in a very short time.
The symptoms of atrial fib can include any of the following: feelings of a fast, fluttering or pounding heartbeat, called palpitations; chest pain; dizziness; fatigue; lightheadedness; reduced ability to exercise; shortness of breath; and weakness.
Current available treatments focus on controlling these symptoms and avoiding dangerous complications, rather than on targeting the molecular origins of the arrhythmia. These include blood thinners to prevent clots that could lead to stroke and surgery to stop faulty electrical signals in the heart. Dr Ellinor noted that employing an invasive surgical procedure reflects the unfortunate fact that therapies that address the underlying causes of atrial fib are not currently available.
One of the procedures currently used to treat atrial fib is percutaneous coronary intervention (PCI)—a non-surgical procedure to treat blockages in a coronary artery and restore healthy blood flow to the heart. In a PCI procedure, a small balloon is inserted into a narrowed or blocked artery to expand and improve blood flow. In many PCI procedures, a mesh wire tube, called a cardiac stent, is inserted and expanded inside a blocked artery to strengthen and support the blood vessel. The stent usually contains medication that releases directly into the artery (drug-eluting stent) to reduce the risk of re-narrowing within the stent.
Over the past two decades, researchers have conducted genome-wide association studies to identify common DNA changes that raise the risk for developing atrial fib. Those efforts yielded more than 140 genetic regions linked to atrial fib risk, but it was clear there were more to find. The Broad Institute has conducted two studies which have greatly added to the number of genetic variants that boost the risk for atrial fib.
In one project, researchers analyzed results from dozens of large genetic studies and uncovered more than 350 common DNA variants associated with risk, doubling the number of common genetic risk factors for the condition. In another, scientists analyzed genetic sequencing data from thousands of individuals with atrial fib and pinpointed rare changes in several genes, which underscore the genetic links between atrial and structural abnormalities of the heart known as cardiomyopathies. The scientists say some of these genes may be at the root of atrial fib and are potential targets for new drugs. The two studies also provide the most detailed look yet into the genetic architecture of this common arrhythmia.
Researchers gathered data from 68 studies from around the globe involving more than 180,000 individuals with atrial fib and nearly 1.5 million individuals without the condition. This meta-analysis identified more than 350 genomic sites associated with atrial fib, twice as many as had been previously identified. In nearly 140 of these sites, the team found genes involved in muscle cell contraction and communication and also heart muscle development. These genes are also more likely to be expressed in atrial heart muscle cells than other genes. Moreover, the team used a new polygenic score to calculate that these new genes are likely to have a stronger cumulative impact on atrial fib risk than previously discovered genes.
In the a second study, Dr Ellinor and colleagues took advantage of recently released genome sequencing datasets to explore uncommon variants that might have sizable impacts on atrial fib risk. Compared to common DNA variants, which may only point to the genome region where the inconsequential DNA misspellings are found, rare variants are more likely to be the specific DNA change that directly leads to cellular dysfunction. “Misspellings,” by the way, is the term used to describe DNA sequences where the nucleotide sequences are accidentally and randomly jumbled, so that the DNA does not function correctly.
Here’s where our description of this research gets into the complex details.
Dr Ellinor and his colleagues gathered whole-genome and whole-exome sequencing data from over 50,000 individuals with atrial fib and more than 270,000 without, and discovered genetic misspellings in four genes (MYBPC3, LMNA, PKP2, and KDM5B ) never before linked to atrial fib. (The exome, by the way, is a very small part of the gene, but is contains most of the disease-related variants, so information about the exome is quite valuable.)
They also observed large effects on risk from deletions in the CTNNA3 gene and from duplications – meaning extra bits of DNA – in the GATA4 gene. Some of these genes are also well known for their role in inherited structural heart defects, pointing to a shared biological basis with atrial fib. To explore the effects of one of these genetic changes, the team used gene editing to turn off the KDM5B gene in stem-cell-derived atrial heart muscle cells, revealing the gene’s involvement in electrical activity in the heart’s atrium, a key process that goes awry in atrial fib.
The researchers are now working to assess any prognostic implications of the results, such as impacts on heart disease risk in individuals carrying the variants. They are also conducting functional studies to attempt to uncover the mechanisms in cardiac activity affected by these genetic variants.
As with the two projects described above, the atrial fib studies are examples of the general trend in healthcare research, which is to pinpoint exact causes and identify precise mechanisms, so as to be able to address those causes and correct their consequences to our health. Think of it as refining the tools: we don’t need a sledgehammer to push a tiny tack into a piece of wood.
The prostate cancer outlook
We have previously discussed prostate cancer several times in these dispatches, but it’s nonetheless worthwhile to take another look at the present status.
Here is some background data from the American Cancer Society:
- Prostate cancer is the second-leading cause of cancer death in American men (lung cancer is the number one cause of cancer death in American men). One in 44 American men will die of prostate cancer. This cancer is the cause of 5.8% of all cancer deaths.
- After skin cancer, prostate cancer is the most common cancer in men in the US.
- Estimated incidences of prostate cancer in 2025: about 313,780 new cases; about 35,770 deaths.
- About 1 in 8 men will be diagnosed with prostate cancer during their lifetime. But each man’s risk of prostate cancer can vary, based on his age, race/ ethnicity, and other factors.
- For example, prostate cancer is more likely to develop in older men. About 6 in 10 prostate cancers are diagnosed in men who are 65 or older, and it is rare in men under 40. The average age of men when they are first diagnosed is about 67.
- Prostate cancer risk is also higher in African American men and in Caribbean men of African ancestry than in men of other races.
- Prostate cancer can be a serious disease, but most men diagnosed with prostate cancer do not die from it. In fact, more than 3.3 million men in the United States who have been diagnosed with prostate cancer at some point are still alive today.
Blood tests for prostate cancer began to emerge in the 1980s, and they have been controversial from the start. Due to the availability of tests, the reported incidence of cases of prostate cancer increases – that’s the reported incidence. That doesn’t mean that the actual incidence increased. In 1985, the estimated lifetime risk of a diagnosis of prostate cancer was 9%. By 2009, this statistic had increased to 16%. This does not mean that more men were actually developing prostate cancer, but that – because the prostate-specific antigen (PSA) tests were becoming increasingly available – more men were being diagnosed with prostate cancer.
Prostate cancer mortality rates declined considerably, at the rate of 2.6% per year, from 2004 to 2012, as more men had prostate-specific antigen (PSA) tests during that time span, which reliably predict the risk of prostate cancer. PSA screening rates during that period were mostly unrelated to the recommendations of the US Preventive Services Task Force (USPSTF). PSA tests were available, and doctors recommended them to their patients.
In 2008, the USPSTF instituted what they termed a Grade D recommendation for PSA screening for men over age 75. Here’s the text of that recommendation as shown on their current website:
“The USPSTF concludes that for men younger than age 75 years, the benefits of screening for prostate cancer are uncertain and the balance of benefits and harms cannot be determined. For men 75 years or older, there is moderate certainty that the harms of screening for prostate cancer outweigh the benefits.”
In 2012, however, the USPSTF specifically stopped recommending annual PSA screenings, whereupon the mortality rates plateaued. Then, in 2018 the USPSTF began recommending that men between the ages of 55 and 69 discuss “possible benefits and harms of screening with their doctors.” According to the USPSTF, the guidelines were changed in order to reduce the number of prostate cancer patients who were treated with “potentially harmful interventions” for non-threatening forms of the disease.
As far as I can tell, the USPSTF has made no changes to their 2018 recommendations.
Probably because of the changes in the USPSTF recommendations, the prostate cancer incidence rate has increased since 2012 by about 3% per year. The USPSTF’s warnings about “possible benefits and harms of screening” and “potentially harmful interventions” refer to the procedures that follow up a positive PSA test result. If the blood test detects the prostate-specific antigen, the usual follow-up is a biopsy of the prostate gland itself. This is indeed a somewhat invasive procedure, usually done by inserting the needle through the walls of the rectum. If cancer cells are found, they are graded according to what is called the Gleason score, which rates the cancer on a scale from 2 to 10 – nonaggressive cancer to very aggressive cancer. Based on the Gleason score, the physician may recommend surgical removal of the prostate.
Now a study by the University of California San Francisco (USCF) reports that the prostate cancer incidence rate increased 6.7% per year between 2011 and 2021. The study included nearly 388,000 men who had prostate cancer between 2004 and 2021. Although the number of cases rose, prostate cancer mortality rates declined 2.6% per year from 2004 to 2012, and plateaued from 2012 to 2021. These trends were similar across age, race, ethnicity and region, the researchers found.
These changes in the prostate cancer mortality rate may be, at least in part, a consequence of the changes in the USPSTF recommendations. An article in JAMA has this brief take on that possible relationship:
“Was the 2012 US Preventive Services Task Force (USPSTF) Grade D recommendation against prostate-specific antigen (PSA) screening for all men associated with prostate cancer–specific mortality (PCSM)?
This cross-sectional study found statistically significant changes in PCSM rates that coincided with the change in the screening guideline; PCSM rates were decreasing prior to the recommendation and remained steady after the recommendation.” (JAMA Netw Open. 2022;5(5):e2211869. doi:10.1001/ jamanetworkopen.2022.11869)
The authorities – JAMA, the University of California, and others – seem reluctant to pin the blame for the reversal of progress in preventing prostate cancer mortality squarely on the USPSTF. In my view, the answer to the question posted in the JAMA article is a loud “YES!”
In contrast with the USPSTF recommendations, the American Urological Association’s guidelines start testing earlier. Here, in part, is the text of their guidelines:
“Clinicians may begin prostate cancer screening and offer a baseline PSA test to people between ages 45 to 50 years. (Conditional Recommendation; Evidence Level: Grade B)
Clinicians should offer prostate cancer screening beginning at age 40 to 45 years for people at increased risk of developing prostate cancer based on the following factors: Black ancestry, germline mutations, and strong family history of prostate cancer. (Strong Recommendation; Evidence Level: Grade B)
Clinicians should offer regular prostate cancer screening every 2 to 4 years to people aged 50 to 69 years. (Strong Recommendation; Evidence Level: Grade A)”
The USPSTF, by the way, is not an official US government agency, nor does the “task force” consist only of health professionals. A substantial part of their focus is on the economics of healthcare. Up until April of 2024, as you may remember, the USPSTF recommended mammograms for women every two years starting at age 50. Then, just about a year ago, responding to considerable criticism, they changed their recommendation regarding the age at which women should start having mammograms, every two years, from age 50 to age 40. In contrast, the American Cancer Society recommends annual mammograms for most women starting at age 45, and annual mammograms starting at age 25 for women considered to be at elevated risk.
On any given healthcare issue, the USPSTF is likely to take a “go slow” approach, mostly in the interest of economy. Yes, there can be harms associated with excessive screening and diagnostic procedures, but the harms associated with failure to detect a serious medical threat should not be underestimated. Regular PSA tests are simple, inexpensive, and effective, and should not be skimped.
The role of artificial intelligence in healthcare
I have read several discussions about AI in healthcare, and I had hoped to be able to summarize these discussions and come to some kind of valid conclusion beyond the obvious. That being, yes, in some cases it can be helpful. But the discussions I read were not very enlightening concerning the particular types of cases in which AI would actually be helpful.
For example, there was an article in from Harvard Medical School entitled “Using AI to repurpose existing drugs for treatment of rare diseases,” by Ekaterina Pesheva, suggesting that AI could identify possible therapies for thousands of diseases, including ones with no current treatments. The article notes that there are more than 7,000 rare and undiagnosed diseases globally. Although each disease affects only a few individuals, collectively these diseases take a huge human and economic toll because they affect some 300 million people worldwide. Yet, with only 5% to 7% of these conditions having a drug that is known to work, they remain largely untreated or undertreated.
In response to this situation, an AI model called TxGN has been developed specifically to identify drug candidates for rare diseases and conditions for which no treatments currently exist.
What this tool did, according to that article, was to identify drug candidates from existing medicines for more than 17,000 diseases, for most of which there were no existing treatments. This, they said, represents the largest number of diseases that any single AI model can handle. The researchers note that, with some modification, the model could be applied to even more diseases beyond the 17,000 it worked on in the initial experiments.
Marinka Zitnik, a researcher at the Blavatnik Institute at Harvard Medical School has this to say:
“With this tool we aim to identify new therapies across the disease spectrum but when it comes to rare, ultra rare, and neglected conditions, we foresee this model could help close, or at least narrow, a gap that creates serious health disparities. This is precisely where we see the promise of AI in reducing the global disease burden, in finding new uses for existing drugs, which is also a faster and more cost-effective way to develop therapies than designing new drugs from scratch.”
I did not see any reference to specific cases in which AI had contributed to favorable health outcomes in an individual.
It seems to me that what AI can most usefully do is sort through and analyze colossal quantities of data. An AI tool can go through the Handbook of Organic Chemistry and match every compound with the pathogen the compound could target. AI can analyze the conformation of the compound and determine which pathogenic particulars the compound can latch onto so as to exterminate the pathogen. AI can also analyze cells and micro-organisms and determine which ones are likely to be able to latch onto and affect parts of the human anatomy, including specific cells.
I am uncertain as to whether, based on this information, AI would be able to determine whether these cells and micro-organisms would cause any harm to humans, and whether attacking them would provide benefit or protection.
But what I am quite sure about is that AI doesn’t care a hoot whether those cells/micro-organisms, or, indeed, anything else, brings harm to humans or contributes to their health. “Caring” is beyond the capabilities of AI. AI could identify a potential disease-causing pathogen that could lead to the extinction of the human race and not do anything more than note it. It might, if so programmed, flag it as a threat. But, essentially, AI doesn’t care.
This leads me to my conclusion that, when it comes to human health and welfare, my confidence in AI is extremely limited. Or, to put it in blunter terms, I do not trust AI. Yes, AI can be useful, but when it comes to your own care, you should put your trust in a human MD.
Not to be alarmist, but there certainly appears to be a looming shortage of human doctors. According to new projections published by the Association of American Medical Colleges, the US will face a physician shortage of up to 86,000 physicians by 2036. Reasons cited include “burnout” exacerbated by the COVID pandemic. However, about a fifth of practicing doctors are currently 65 years old or older, and are looking forward to a life without the stresses and demands of providing healthcare to their patients. These demands are on the increase, also because of the increase in the age of the general population, which increases the demands on the healthcare system.
A factor that is cited is that Medicare has failed to increase its funding of medical school scholarships sufficiently to meet the demand for physicians. Personally, I question the impact of that factor. Young people were attracted to the medical profession for many reasons – doctors made a pretty good living, were well respected in their community, and did really interesting work. What has changed?
For one thing, the healthcare system itself has changed in a way that makes physicians spend a huge amount of time staring at their computer screens, entering data demanded by health insurance plans, Medicare, Medicaid, or whatever. And when they look to their computer for information, some of that information may be supplied by AI. Healthcare professionals are overworked, and may not take the extra time to go beyond the quick answer that AI gives them. And AI relies on “information” that is on the internet, some of which may be incorrect.
The combination of increasing reliance on AI and the looming shortage of healthcare workers certainly poses a highly concerning problem to individual patient care. Speaking for myself and my spouse, we are exceedingly fortunate to have established long-term relationships with excellent primary care physicians and other specialists. Both of our PCPs keep paper files and interact with us directly, without staring into their computer screens. When we call their offices, a human being answers the phone, and this person actually knows who we are! As I said, we are very fortunate, and we hope things stay this way!
* * * * * * * *
I had promised, in a previous missive, to take a look at digestive tract diseases – Crohns, ulcerative colitis, and others – and see what progress has been made in managing these threats. But, as you see, I was overwhelmed by releases from several sources than came flooding into my inbox and demanded my attention. More releases have showed up in my inbox, and more will certainly appear, but I will for sure look at those diseases in my next dispatch.
Looking forward to a real Spring! Be well, everybody! Best, Michael Jorrin (aka Doc Gumshoe)
[ed note: Michael Jorrin, who I dubbed “Doc Gumshoe” many years ago, is a longtime medical writer (not a doctor) and shares his commentary with Gumshoe readers once or twice a month. He does not generally write about the investment prospects of topics he covers, but has agreed to our trading restrictions. Past Doc Gumshoe columns are available here.]
