Dr. Del Tredici reviews 5 of her favorite papers from 2019. We discuss aortic stenosis, the SGLT2 inhibitors, the safety of PPIs, and whether or not you can treat osteomyelitis with oral antibiotics.

Presenter: Sonya Del Tredici, MD
Host: Giselle aerni, MD
Producer: Robert Stuntz, MD

PAPER 1: TAVR and more TAVR

Transcatheter Aortic-Valve Replacement with a Balloon-Expandable Valve in Low-Risk Patients MJ Mack, et. al. The PARTNER-3 Investigators. The New England Journal of Medicine, 380 (18): 1695-1705. May 2, 2019.

Why is this paper important?

More TARVs than SAVRs are performed in the US, but they are traditionally reserved for intermediate and high-risk patients. This trial aimed to evaluate TAVR in low-risk surgical patients with aortic stenosis.

What is the clinical question?

Population: Patients with severe calcific aortic stenosis and low surgical risk. They excluded patients with low life expectancy, bicuspid aortic valves and disqualifying anatomy. The study included mostly American men, age ~71, 28% with CAD, 30% with DM, 15% with afib.

Intervention: Transfemoral TAVR with a balloon-expandable valve, along with ASA+clopidogrel.

Comparison: Surgical aortic valve replacement with a bioprosthetic.

Outcomes: Patients were followed for 1 year.
Primary endpoint: composite of all-cause mortality, stroke, and rehospitalization related to the valve.
Secondary endpoints: new-onset afib, length of stay, death (30 d and 1 yr), stroke (30 d and 1 yr), rehospitalization, overall bad outcomes.
These outcomes were both clinical and physiological.

Is the study valid?

I think this was a well-done study, and the conclusions were valid. It was randomized, but non-blinded. They enrolled 1000 patients, which was enough to power the study. The patients were similar at baseline. They had an appropriate intention-to-treat analysis, and pretty complete outcomes data. About 10% of patients in the surgery group withdrew, deciding they didn’t want surgery. The subgroups were pre-specified.

Adverse events were reported.

Some limitations were that they only collected 1 year of data, and since the patients were relatively young and healthy, they have a lot more than 1 year to go. I would have preferred to see what happened to them over a longer time. It was funded by the manufacturer of the valve, which may have caused some bias as well.

What are the results?

The results showed that TAVR is safer, has better outcomes at 30 days, and has better outcomes at 1 year. The patients who got a TAVR had shorter length of stay, fewer operative complications, were more likely to be discharged to home. They also had less afib, stroke, major bleeding, rehospitalization, and death. The only thing that was worse for the TAVR group was incidence of new LBBB.

How will this study help us in patient care?

The patients in this study are similar to my patients. The outcomes were clinically relevant. The benefits were both clinically and statistically significant. And TAVR is both cheaper and easier for patients. Therefore I conclude that this paper establishes TAVR as the treatment of choice for most patients whose anatomy allows it, especially older patients unconcerned about long-term valve durability. From now on surgical risk should no longer determine who should have a TAVR. Instead we can now look at life expectancy, anticoagulation needs, and long-term valve durability.

PAPER 2: Doc, I feel fine. Do I have to get my valve replaced?

Early Surgery or Conservative Care for Asymptomatic Aortic Stenosis The RECOVERY Trial. D Kang, S Park, S Lee, Et. al. N Eng J Med 382 (2), 111-119, January 9th, 2020

Why is this paper important?

In asymptomatic patients with aortic stenosis, the ideal time for surgical intervention is not known. Because the surgery is high-risk, previous guidelines recommended only doing valve replacement on symptomatic patients. But asymptomatic patients can die from sudden death, and they can sustain irreversible myocardial damage, causing later morbidity. But now that surgical techniques have improved, perhaps the risk-benefit calculation has changed, and we should be doing valve replacement on patients before they become symptomatic.

What is the clinical question?

Population: Included: adults with severe AS.
The demographics of the participants were Korean adults, enrolled 2010-2015, mean age 64, BMI 24, HTN 40%, high chol 40%, bicuspid aortic valve 60% (which is important because that makes TAVR not an option), degenerative valve 33%.
Excluded: symptomatic AS, CHF, other valve problems, and those who were not surgical candidates.

Intervention: SAVR (50% mechanical, 50% biological).

Comparison: conservative care, but offered surgery if AS became symptomatic. 74% eventually got surgery.

Outcomes: primary endpoint: operative mortality (w/in 30 days) or CV death within trial period (4-7 yrs)
Secondary endpoint: mortality, stroke, MI, repeat AV surgery, CHF hospitalization

Is the study valid?

It is randomized, not blinded, multi-center study. There were 145 patients enrolled, which gave the study sufficient power. The two groups were similar at baseline.

There was an intention-to-treat analysis. No patients lost to follow-up. The funding was by the Korean Institute of medicine, and was unlikely to cause bias.

What are the results?

The results showed that patients who got the early surgery, while they were still asymptomatic, had lower rates of death and CHF. The early surgery group had 7% risk of death. The late surgery group had 21% death. And the early surgery group had no incidence of sudden death, whereas the late surgery group had 14% die of sudden death. The early surgery had somewhat higher rates of stroke: 3% vs. 1 %.

How will this study help us in patient care?

This high-quality study showed that it is better to operate early on severe aortic stenosis, and not wait for symptoms to develop. As well as surgery being inevitable in these patients, delaying surgery put the patients at risk of sudden death. I will be referring all of my patients with severe aortic stenosis for valve replacement.

PAPER 3: Not just for diabetics any more!

Dapagliflozin in Patients with Heart Failure and Reduced Ejection Fraction JJV McMurray, Et. Al. N Eng J Med 381(21), 1995-2008. November 21th, 2019

Why is this paper important?

The SGLT2 inhibitors are a novel class of agents to treat heart failure. SGLT2 inhibitors had been shown to reduce new-onset CHF as a secondary benefit in DM, but had not been investigated in CHF patients without diabetes..

What is the clinical question?

Population: Inclusion: adults, EF<40%, NYHA class II, III, or IV, not on an SGLT2-I.
Exclusion: DM-I, SBP<95mmHg, eGFR<30, intolerance of SGLT-2-I
Demographics: NYHA class II: 67%, class III: 32%, EF ~ 30%, 77% male, age ~66, 70% white and 24% Asian, 42% had DM.

Intervention: dapagliflozin 10 mg daily, as well as all standard CHF therapy

Comparison: Placebo

Outcome: Primary: composite of CHF hospitalization, IV diuretics, and CV death.
Secondary: CHF hospitalization, death, decrease in the KCCQ score, worsening renal function, renal death, all-cause mortality.
These outcomes were clinical and relevant.

Is the study valid?

This is a valid study. It was randomized (stratified for diabetes), Placebo-controlled, and blinded. It included enough patients to adequately power the study (4744). The groups were similar at baseline. The duration was a median of 18.2 months, which is long enough to see some outcomes. There was an intention to treat analysis. 11% stopped the medications for a variety of reasons, but it was equal in intervention and control groups. Adverse events were reported, including volume depletion, declining renal function, hypoglycemia, fractures, DKA, and GU infections.

The only limitation that concerned me was that it was funded by AstraZenica, which may have caused some bias.

What are the results?

The study showed that the patients on dapagliflozen had better outcomes than the patients taking placebo. They had fewer CHF exacerbations and death, and as an added bonus their diabetes was better controlled. The adverse events were the same in the two groups as well, except for renal adverse events, which were better on dapagliflozen.

How will this study help us in patient care?

The benefits of dapagliflozen in CHF were both clinically and statistically significant. These patients were sicker than those in other trials of SGLT-2 inhibitors, and were already receiving maximum CHF therapy, including loop diuretics and mineralocorticoid receptor antagonists. This improved outcomes even beyond that, and no increased adverse effects were noted. The only downside is the cost: about $500 per month retail. So I will be adding an SGLT2-inhib to my patients with CHF, if they are still facing exacerbations on maximal medical therapy.

PAPER 4: Cum hoc, ergo propter hoc (with this, therefore because of this. Correlation does not imply causation)

Safety of Proton Pump Inhibitors Based on a Large, Multi-Year Randomized Trial of Patients Receiving Rivaroxaban or Aspirin. P. Moayyedi, Et. Al. Gastroenterology 157, 682-691. September 2019

Why is this paper important?

PPIs are one of the most widely-prescribed classes of prescription drugs in the US, used for treatment of GERD and a variety of other acid-related conditions. Observational studies have associated long-term PPI use with many diverse adverse outcomes: CV events, PNA, fractures, enteric infections, C. diff, CVA, CKD, dementia and death. The mechanisms are biologically plausible but unproven, and the associations have never been satisfactorily evaluated in a randomized trial. This is the first large-scale prospective randomized trial to look at PPI safety in a population with significant underlying comorbidities. This trial was also evaluating rivaroxiban and aspirin for secondary prevention of CVD, which made it a little confusing. I’m just going to discuss the PPI arm here.

What is the clinical question?

Population: Inclusion: Age≥65 with stable CVD or PAD, or age < 65 with CVD + risk factors.
Exclusion: Need for PPI at baseline, high risk of bleeding, severe CHF, significant CKD, need for dual anti-platelet tx,
Demographics: age ~67, 79% male, 60% white, from Europe and the Americas, 61% had a previous MI38% had diabetes, 25% had CHF22% had CKD

Intervention and Control: Pantoprazole 40 mg daily vs. placebo, followed for ~3.1 yrs

Outcomes: Cardiovascular outcomes: MI, stroke, CV death, CHD, acute limb ischemia
PPI safety outcomes: PNA, c. diff, enteric infections, fractures, gastric atrophy, CKD, DM, COPD, dementia, CVD, cancer, hospitalizations, death

Is the study valid?

This was a valid study. It was blinded, randomized, placebo-controlled, and multicenter. It included 17598 participants, large enough to adequately power the study. The follow-up was 3.1 years, which I think was long enough to see outcomes in this high-risk population. The two groups were similar at baseline. There was an intention-to-treat analysis.

There were some limitations. It excluded those taking PPI at baseline, so may not be a true representative sample of our population of CVD. Also, 22% stopped the medication (placebo and pantoprazole), although it was the same number in both the pantoprazole and control groups. PPI safety data was collected via biannual interviews, rather than review of medical records, and received no independent adjudication. So the data is not as strong as it could have been.

What are the results?

In the patients randomized to pantoprazole 40 mg daily, there was no increase in MI, stroke, death, enteric infections, gastric atrophy, c. diff, CKD, dementia, pneumonia, or fracture.

How will this study help us in patient care?

To summarize, in high-risk patients who were not on a PPI, being randomized to get pantoprazole 4 mg daily for about 3 years did not increase the risk of any of the possible adverse events we attribute to PPI use. So PPIs are likely safe, and we can use them in patients for whom they’re indicated.

PAPER 5

Oral versus Intravenous Antibiotics for Bone and Joint Infection (OVIVA) HK Li, R Romback, Et. Al. N Eng J Med 380(5), 425-436. January 30th, 2019

Why is this paper important?

Treatment of complex orthopedic infections is difficult, expensive, and risky to the patient. There is a general feeling (unsupported by evidence) that antibiotics by IV are superior to PO. A small meta-analysis in 2012 showed no advantage to IV over PO, but it was not quite big enough to change practice. This paper aimed to determine if oral antibiotics were as good as IV antibiotics.

What is the clinical question?

Population: Inclusion: age>18, requiring 6 weeks abx for bone or joint infection (osteo, joint infx with arthroplasty, prosthetic joint infx, orthopedic device infx, vertebral osteo)
Exclusion: Could not be randomized due to a need for either PO or IV Abx
Demographics: UK, av age 60, 64% male

Intervention: 6 weeks of oral antibiotics

Comparison: 6 weeks of IV antibiotics

Outcomes: Primary: definite treatment failure at 1 year (clinical, microbiological, or histological)
Secondary: possible treatment failure at 1 yr, discontinuation of antibiotics, catheter complications, c. diff, resource use, health status, and adherence.

Is the study valid?

This was a valid study. It was randomized, pragmatic, multicenter, but non-blinded. Sample size was large enough, with 1054 participants, to show non-inferiority.

There was an intention-to-treat analysis, and endpoint data available for 96% of participants. The one-year follow-up was long enough. The two groups were similar at baseline. The only difference between then was that the IV group may have received more frequent assessments and more overall care than the PO group.

What are the results?

This study showed that oral antibiotics were as good as IV antibiotics for treating infections, and had the same number of adverse outcomes. Both groups had about 14% treatment failure, and about 27% had some kind of adverse event.

How will this study help us in patient care?

This study challenges a widely held standard of care, which is that you need 6 weeks of IV antibiotics to treat complex bone and joint infections. But oral therapy was a good as IV along all parameters tested, and so should become the new standard of care.

This episode was recorded on 5/7/2020.