COMMENTARY

October 29, 2021 This Week in Cardiology Podcast

John M. Mandrola, MD

Disclosures

October 29, 2021

Please note that the text below is not a full transcript and has not been copyedited. For more insight and commentary on these stories, subscribe to the This Week in Cardiology podcast on Apple Podcasts, Spotify, or your preferred podcast provider.

In This Week’s Podcast

For the week ending October 29, 2021, John Mandrola, MD comments on the following news and features stories.

AC in Post-stroke Patients With LV Dysfunction

How best to treat a patient with a stroke of unknown source comes up often, especially for electrophysiologists, who are often asked to place ILRs (loop recorders) to monitor for atrial fibrillation (AF). JAMA-Neurology has published an interesting post-hoc analysis of the NAVIGATE ESUS trial of rivaroxaban vs aspirin (ASA) in patients who have had an embolic stroke of unknown source (ESUS).

The question of this analysis was whether left ventricular (LV) systolic dysfunction modified the effect of rivaroxaban vs ASA—was there heterogeneity of treatment effect. It’s a good question because impaired systolic function can increase the risk of thrombus and thrombotic complications.

The New England Journal of Medicine (NEJM) published the original NAVIGATE ESUS trial in 2018. It randomly assigned about 7200 patients with ESUS to rivaroxaban 15 mg vs ASA 100 mg and found no difference in recurrent stroke or systemic embolism. Major bleeding was 2.7 times more likely in the rivaroxaban group. (It wasn’t only rivaroxaban. Dabigatran also failed to significantly reduce stroke in the RE-SPECT-ESUS trial of standard dose dabigatran vs ASA in patients with ESUS. In this trial, dabigatran reduced strokes by 15%, but the P-value came in at 0.10.) These results always seemed weird to me. If you have an embolic stroke and no large artery atherosclerosis, it stands to reason oral anticoagulation (AC) would be better. You wonder if it’s patient selection.

The authors of the JAMA-Neurology paper looked at the patients in NAVIGATE-ESUS who had abnormal LV systolic function On baseline echocardiogram—7% of the 7200 total, or 500 patients.

  • Among these 500 patients, a recurrent stroke or systemic embolism occurred in 2.4% of those in the rivaroxaban group vs 6.5% in those assigned to ASA.

  • Among the 6600 patients in the trial who had normal LV systolic function, the rates were 5.3% vs 4.5%.

  • Participants with LV dysfunction assigned to rivaroxaban vs ASA had a 64% lower risk of stroke (hazard ratio [HR] 0.36; 95% CI, 0.14-0.93)—clearly significant.

  • Those without LV dysfunction showed no such reduction (HR, 1.16; 95% CI, 0.93-1.46). The test for an interaction was significant at 0.03.

Comments. Does this mean we should treat ESUS patients with impaired systolic function with oral AC? No, it does not, and the authors are clear on this point, both in the manuscript and in the news coverage.

Studies like this are essentially observational comparisons, and if you do enough post-hoc comparisons, some will be significant based solely on chance. A chance finding is entirely possible in this comparison because the delta was based on only 22 strokes, six in the rivaroxaban group and 16 in the ASA group. But this is a biologically plausible association. We have all seen embolic complications in patients with impaired LVs.

If you go to the paper, you will find a very nice comment from the esteemed neurologist David Spence. He made the point that this is now the second subgroup analysis of NAVIGATE -ESUS to show lower rates of stroke with rivaroxaban over ASA. The first was a subgroup analysis looking at ESUS patients who had patent foramen ovale (PFO). Spence points out that the likely reason ESUS trials fail to show superiority of oral AC over ASA is misclassification of large artery atherosclerosis—carotid disease, for example. Obviously, if you Include patients with carotid disease, you favor the trial towards ASA. In the ESUS trials, Spence writes, large artery disease is defined by the presence of stenosis. But due to compensatory enlargement, many patients with high plaque burden do not have high-grade stenosis.

So, I wonder, if we better classified carotid disease in ESUS patients, excluding patients with carotid plaque burden, and/or included those with LV dysfunction, we might be able to show that AC is superior to ASA. The specific message here is that maybe there are groups of ESUS patients in whom oral AC may be of benefit.

But the general message of this story is that nonsignificant trials can still teach us a lot. Trials with nonsignificant primary endpoints shouldn’t tempt us to think binary—that is, this doesn’t work. But rather, maybe it does work in selected patients.

The point of post-hoc subgroup analyses is not to mistakenly make causal claims, but to inform future trials, such as a trial of ESUS patients with either a PFO or LV dysfunction and limited carotid plaque burden. Of course, the medical conservative approach to medicine is that the onus of proof is on proponents of a therapy to show it works in some group of patients.

Coronary Artery Calcium Scores

Coronary artery calcium score (CAC) was back in the news this week. But before I discuss CAC, a disclosure: I am on record as being a skeptic that CAC has any role. Period. Andrew Foy and I published our argument in American Family Physician. We are especially worried about the CAC 0 argument, which holds that CAC 0 confers such a low future risk of events, that it can allow stopping statin therapy. The problem with that logic is that statins have been shown in oodles of RCTs, to lower future event rates by about 25%. And CAC 0 has been studied only in observational studies. So, by saying you can use CAC 0 to stop statins, you are using observational studies to refute RCT-level data.

JAMA-Cardiology published a cohort study of roughly 24,000 patients with known obstructive coronary artery disease (CAD) to determine the diagnostic value of a CAC score of 0 to rule out obstructive CAD. The authors used a Danish registry and included patients who underwent computed tomography angiography (CTA) because of symptoms. The median follow up was about 4 years.

There are four main findings:

  • The authors first tell us the prevalence of obstructive disease in those with CAC 0—6% (725 of 12,771), ranging from 3% in those younger than 40 to 8% in those over 70.

  • The authors then tell us the prevalence of CAC 0 in patients with obstructive disease—14% overall (725 of 5043). Crucially this varied by age:

    • Younger adults with obstructive disease were far less likely than older adults to have any CAC.

    • 58% of patients (39 of 68) younger than 40 with obstructive disease had a 0 CAC score compared with only 5% of patients (52 of 964) who were 70 years or older with obstructive disease.

    • This shouldn’t be surprising because plaques are less likely to be calcified in younger people.

  • The authors then tell us the diagnostic value of a CAC of 0 for ruling out obstructive disease, adjusting for other risk factors and present a DLR (diagnostic likelihood ratio). Higher values close to 1 mean the CAC 0 modifies the likelihood less, while lower values modify the risk more. Again, age mattered:

    • In those younger than 40, the DLR of CAC 0 was 0.68 or just 32% less likely to have obstructive disease.

    • In those older than 70, the DLR of CAC 0 was 0.18, meaning these patients were 82% less likely to have obstructive disease.

Figure 2 in the paper presented an example of a nonsmoker male without diabetes and atypical chest pain, with and without CAC. In those younger than 40 years, the risk of obstructive disease using clinical variables was 6%, and if the CAC was 0, the risk was now 4%. A delta of only 2%.

But for the same nonsmoking, nondiabetic male with atypical pain who was older than 70 years, the risk of obstructive disease using clinical variables was 40% but a CAC 0 drove it down to 8%, a 32% absolute risk reduction.

  • The final data-point the authors provided was the rate of myocardial infarction (MI) and death based on CAC of 0.

  • Among those with a CAC of 0 and obstructive disease, the rate of MI or death was 7 per 1000 patient years compared with 4 per 1000 patient years in those with CAC of 0 and no obstructive disease.

  • The HR for obstructive vs non-obstructive disease in CAC 0 patients was 1.5—not surprising that obstructive disease raised MI and death risk.

  • What was novel was that HR for obstructive vs non-obstructive disease in CAC 0 also turned-on age: In those with CAC 0, younger patients had an 80% higher risk of an MI or death if they had obstructive disease whereas older patients only had a nonsignificant 24% higher risk.

The authors concluded: The diagnostic value of a CAC score of 0 to rule out obstructive CAD beyond clinical variables depended on age. The added diagnostic value of CAC 0 was smaller for younger patients.

In symptomatic patients who were younger than 60 years, a sizable proportion of obstructive CAD occurred among those without CAC and was associated with an increased risk of myocardial infarction and all-cause death.

The accompanying editorialists wrote: The article by Mortensen et al  should give pause to efforts to broaden the use of a CAC score of 0 to de-escalate or defer statin therapy in all individuals.

I totally agree. We know that atherosclerosis begins early and worsens over years. Calcification takes time to develop. If you measure a CAC of 0 in a 40-year-old and use it to stop statins, you stand a not-small chance of missing obstructive disease and taking away an intervention that reduces events. If you find a CAC 0 in an older person, who has had time to develop calcium, okay, you are less likely miss obstructive disease.

This study supports Foy’s and my worries about de-escalating statins based on observational data—especially in younger people. For instance, we know that length of time exposed to a lower LDL reduces the risk of events. That means the benefit of statins accrues over time; the young surely have more to gain from preventive therapies than the old. Preventive therapies are in fact designed to create old people.

As the editorialists write: This analysis serves as e a reminder to physicians that the absence of CAC is not equivalent to the absence of atherosclerosis, particularly in younger adults and women.

Finally, there would be no temptation to remove beneficial therapies if there were no CAC scores. I remain a massive skeptic of this radiologic test.

Takotsubo Syndrome

The Journal of the American Heart Association published a research letter on the sex- and age-based trends in Takotsubo cardiomyopathy. Takotsubo syndrome (TTS), also known as broken heart syndrome or stress-induced cardiomyopathy, is an intriguing condition that mimics a ST-segment elevation MI (STEMI). Patients can get very ill and then improve over days to weeks. We see this condition often.

The researchers used the National Inpatient Sample (NIS) database to report on trends from 2006-2017.

  • The annual incidence in TTS increased steadily in both sexes with women comprising 88% of cases, especially women older than 55 years.

  • The most prominent at-risk group is women between ages 50-74 years.

The authors speculate on why there is such an increase in women, but it’s pure speculation. While we know sympathetic stimulation is related, every human being has periods of sympathetic stimulation and doesn’t get the syndrome.

I mention this paper for a couple of reasons: one is that this is a good use of observational data, telling us about trends over time. Second, I find everything about this condition intensely curious. Why does it occur? Why in middle-aged women so predominantly? Why is it increasing? What are we missing in the environment?

If you have ideas, let me know in the comments.

Chest Pain Guidelines

Numerous authors have published an exhaustive guideline document for the evaluation and treatment of patients with chest pain. The executive summary is 44 pages with 350 references.

First, congratulations to the authors, especially chairperson Martha Gulati. Documents like these are a massive undertaking. And the text and references will serve as an incredible reference and teaching tool. Such is a worthy endeavor, one that improves the greater good.

As with all massive new guideline documents, the authors give us the top-ten take home messages. I understand why this is done, but I don’t like it. It encourages folks to read the take-homes and not delve into the nuance.

Along those lines, the authors also put tons of stuff into neat little colored boxes with strength of recommendation and level of evidence. This is standard practice for guidelines, but I don’t like it. Colored boxes belie the complexity of medicine. They simplify things that really cannot be easily simplified. Chest pain evaluation, especially. The main ingredient for success in treating patients with chest pain is solid clinical skills along with extreme powers of listening and observation. A robot will fail miserably at chest pain evaluation.

Most of the recommendations for in this guideline come with low level evidence and thus lend themselves to the narrative reviews, which are superb. I have long held that all guidelines would be improved if only narrative review and references were included.

Some specific points on the document:

  • The authors emphasize accompanying symptoms. While pain in the chest is common, it’s crucial to pay attention to things like dyspnea, arm pain, and jaw pain.

    • Always be alert when someone says jaw pain.

    • Always look at people. People having unstable angina and MI often look unwell.

  • I love the authors’ choice to make shared decision making the third point of emphasis in the 10 take-homes. A clear discussion of the uncertainties and possibility of adverse events, radiation exposure, costs, and alternative options is a positive on its own.

    • Shared decision making is correct care, not because it reduces over testing but because it increases the odds of aligning care with a patients’ goals.

    • Some patients like the idea of more testing, others are fine with risk stratification. Shared decision making should never be used as cudgel to do less.

  • The authors put emphasis on pathways and structured assessment of risk. I understand why they do that, but the thing with seeing chest pain patients are the tails. No one wants to miss an MI or pulmonary embolism (PE). Patients with low heart risk scores can have MIs. And patients with high-risk scores can have noncardiac chest pain.

  • Structured assessments and risk scores are good for the middle-of-the-curve patients—but these are the easy ones. I am not convinced that risk scores help that much.

  • As for over-testing and overtreatment, I see only one antidote: someone writes in clear language how many missed PEs and MIs are allowed. For if it is zero, over-testing will continue. You might as well stop fighting against low-value chest CTs and coronary CTAs.

Given the extremely high prevalence of chest pain, I am sure this document will serve as an important reference for many different specialties—internal medicine, family practice, emergency medicine, radiology, and cardiology alike. Good on the authors.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....