Use of a Pulmonary Artery Pressure Sensor to Manage Patients With Left Ventricular Assist Devices

Background:

Hemodynamic-guided management with a pulmonary artery pressure sensor (CardioMEMS) is effective in reducing heart failure hospitalization in patients with chronic heart failure. This study aims to determine the feasibility and clinical utility of the CardioMEMS heart failure system to manage patients supported with left ventricular assist devices (LVADs).

Methods:

In this multicenter prospective study, we followed patients with HeartMate II (n=52) or HeartMate 3 (n=49) LVADs and with CardioMEMS PA Sensors and measured pulmonary artery pressure, 6-minute walk distance, quality of life (EQ-5D-5 L scores), and heart failure hospitalization rates through 6 months. Patients were stratified as responders (R) and nonresponders to reductions in pulmonary artery diastolic pressure (PAD).

Results:

There were significant reductions in PAD from baseline to 6 months in R (21.5–16.5 mm Hg; P<0.001), compared with an increase in NR (18.0–20.3; P=0.002), and there was a significant increase in 6-minute walk distance among R (266 versus 322 meters; P=0.025) compared with no change in nonresponder. Patients who maintained PAD <20 compared with PAD ≥20 mm Hg for more than half the time throughout the study (averaging 15.6 versus 23.3 mm Hg) had a statistically significant lower rate of heart failure hospitalization (12.0% versus 38.9%; P=0.005).

Conclusions:

Patients with LVAD managed with CardioMEMS with a significant reduction in PAD at 6 months showed improvements in 6-minute walk distance. Maintaining PAD <20 mm Hg was associated with fewer heart failure hospitalizations. Hemodynamic-guided management of patients with LVAD with CardioMEMS is feasible and may result in functional and clinical benefits. Prospective evaluation of ambulatory hemodynamic management in patients with LVAD is warranted.

Registration:

URL: https://www.clinicaltrials.gov; Unique identifier: NCT03247829.

What is New?

In the INTELLECT 2-HF study (Investigation to Optimize Hemodynamic Management of HeartMate II Left Ventricular Assist Device Patients Using the CardioMEMS Pulmonary Artery Pressure Sensor in Advanced Heart Failure) of patients implanted with a durable LVAD and a pulmonary artery pressure (PAP) sensor, a significant reduction in PAP over 6 months was associated with improvements in 6-minute walk distance. Furthermore, patients who maintained a diastolic PAP <20 versus PAP ≥20 mm Hg throughout the follow-up period had fewer HF hospitalizations.

What are the Clinical Implications?

Some LVAD patients have residual heart failure post implant and may benefit from tailored management with diuretics and vasodilators or adjustments of LVAD rotational speed. The INTELLECT 2-HF study suggests that hemodynamic-guided management of LVAD patients with an implantable PAP sensor is feasible and may improve functional status and clinical outcomes. Prospective study of this approach is warranted.

Continuous-flow left ventricular assist devices (LVADs) improve survival and quality of life in patients with advanced heart failure (HF).^1–3 Durable LVADs restore systemic perfusion, improve end-organ function, and enhance functional capacity, regardless of whether the therapeutic intent is bridge to transplant or a permanent alternative to transplant. Despite the established benefits of LVADs, up to 20% of patients do not experience improvements in functional capacity and 12% are rehospitalized for heart failure.^4–7

Nearly one-third of patients with LVADs undergoing right heart catheterization require adjustment of LVAD rotational speed to optimize hemodynamics.^8,9 Observational studies and a small, prospective randomized trial have described improved clinical outcomes after hemodynamic ramp studies, suggesting that persistent hemodynamic abnormalities may underlie the lack of functional improvement and recurrent HF events in some LVAD patients.^6,7,10–12

Hemodynamic-guided management using an implantable pulmonary artery pressure (PAP) sensor (CardioMEMS system) is indicated to reduce heart failure hospitalizations (HFH) in patients with symptomatic HF (NYHA class III) and a HFH in the preceding 12 months.^13,14 Reductions of 3 to 5 mm Hg in mean PAP (PAM) were associated with improved functional capacity, reduced hospitalizations, and better overall survival.^13,14 Furthermore, a lack of change in pulmonary artery diastolic (PAD) pressure at 3 months compared with baseline was associated with a greater than 2-fold higher chance (HR, 2.35) of subsequent need for LVAD or heart transplantation,^15,16 suggesting PAP could provide valuable information in the decision-making process for these advanced HF therapies. The use of CardioMEMS to help manage LVAD patients is supported by limited data, but the field is lacking large prospective clinical trials.^17,18

We sought to determine the feasibility and clinical utility of the CardioMEMS HF system to manage patients supported with HeartMate LVADs. Our goal was to evaluate longitudinal PAP changes in continuous-flow LVAD patients and correlate them with clinical end points, including 6-minute walk distance and HFH.

Methods

The INTELLECT 2-HF trial (Investigation to Optimize Hemodynamic Management of HeartMate II Left Ventricular Assist Device Patients Using the CardioMEMS Pulmonary Artery Pressure Sensor in Advanced Heart Failure; ClinicalTrials.gov, NCT03247829) was a prospective, nonrandomized multicenter, single-arm observational study conducted in 22 clinical sites in the United States involving 101 patients. Enrollment lasted approximately 2 years (August 24, 2017 to August 20, 2019) with a 6-month follow-up period ending on March 6, 2020. The Institutional Review Board of each participating center approved the study protocol. Each patient provided written informed consent and agreed to provide access to patient and device data (including CardioMEMS Merlin.net data). The INTELLECT 2-HF trial was executed by a diverse group led by the National Principal Investigators, along with the Podium and Publications Committee, and Sponsor representatives. This group includes representation by sex, age, race, and national origin, which is also represented in the authorship of this article. The complete data supporting the findings of this study are available from the corresponding author upon request.

Patients were enrolled in 2 groups: existing CardioMEMS and new CardioMEMS. Existing CardioMEMS patients had been implanted with either a HeartMate II (HMII) or 3 (HM3) LVAD and a CardioMEMS sensor before enrollment. New CardioMEMS patients had a HeartMate II or 3 LVAD in place and met the indication for CardioMEMS implantation. New CardioMEMS patients had the sensor implanted within 72 hours of consent. Before discharge, subjects were trained to measure PAP using the home electronic monitoring unit. Measurements were made in the supine position and uploaded to Merlin.net. Readings and trend data were reviewed by healthcare professionals experienced in the management of HF at least once every 7 days. Guidelines for hemodynamic-based HF management were supplied as part of the INTELLECT 2-HF protocol, but no specific treatment algorithm was prescribed as the study was designed to observe the interventions used. Protocol guidance was provided to investigators with the goal of reducing PAD by using medications and LVAD speed changes. However, all changes to either medications or LVAD pump speed were at the discretion of investigators and documented changes were based on clinical grounds.

All subjects were followed for 6 months or until withdrawn from the study. Study visits occurred at baseline, 1, 3, and 6 months post-enrollment. Assessments and information for each subject was collected at baseline and at follow-up visits to meet the descriptive end points as detailed in the protocol. Briefly, at baseline, subject medical history, HF assessment, demographic information, device documentation, current medications, pump parameters, as well as 6-minute hall walk test distance (6MWD) and EQ-5D-5 L survey scores, were collected. At each follow-up visit, changes in medications, vital signs, NYHA classification, EQ-5D-5 L survey, 6MWD as well as PAPs before and after the 6-minute hall walk test, HF hospitalizations (HFH) including reoperations/operations/pump replacements/explants/device exchanges pertaining to the devices, and any occurrence of adverse events were assessed.

Further analysis of all patients was performed by stratifying all subjects as responders or nonresponders as well as by time spent in the target range (TTR) outlined in the protocol. Groupings by TTR were identified for those who spent more than 50% of the study follow-up period (6 months) with a PAD greater than or equal to 20 mm Hg (PAD ≥20) compared with those with a PAD <20 mm Hg (PAD<20). A PAD of 20 mm Hg has been previously defined as the upper threshold of normal for the PAD range.^19,20 A responder refers to a patient who has clinically meaningful response to attempts by investigators to reduce PA pressures toward the goal range of PAD <20 mm Hg. For the purpose of this study, responders (R) were defined as subjects who had an average reduction in PAD of at least 1 mm Hg using the area under the curve method (ie, 180 mm Hg*days over 6 months), a threshold based on the mean PAP reduction (ie, 156 mm Hg*days=0.9 mm Hg over 6 months) in the treatment arm of the CHAMPION trial.²¹ Nonresponders had an average PAD reduction of <1 mm Hg or an increase.

Statistical Analysis

Responder, nonresponder, PAD ≥20, and PAD <20 groups were determined using an area under the pressure-time curve of each patient’s daily change in PAD from baseline calculated using the trapezoid rule. Baseline characteristics, including patient demographics, hemodynamics, medical histories as well as device implant sequence and parameters are presented as counts and percentages for categorical variables and mean±SD for continuous data. Baseline categorical variables and continuous data were analyzed using t tests and the Fisher exact test, respectively. Comparisons between 2 group PAPs and 6MW distances within each time point (baseline, 1, 3, and 6 months) were compared using t tests. Longitudinal changes in PAPs and 6MW distances were analyzed using a linear mixed model. Subject HFHs and volume management are presented as percentages and compared using the Fisher exact test. Interventions are presented as counts, percentages, number of events, events per patient year (EPPY), rate ratio, and 95% CI and analyzed using Poisson regression. Repeat interventions were included in the count. Group proportions of patients without any interventions were compared using χ² test. Statistical analyses were performed using SAS software, version 9.4 (SAS Institute, Cary, NC).

Results

Figure 1 depicts the prospectively enrolled cohorts included in the INTELLECT 2-HF study. A total of 101 subjects were enrolled into 2 groups, existing CardioMEMS (n=53: HMII=36, HM3=17) and new CardioMEMS (n=48: HMII=16, HM3=32). Eighty-eight patients completed 6 months of clinical follow-up. Thirteen patients were unable to complete the study due to heart transplantation, device explant, or death. Two subjects did not report baseline and 6-month PADs and are excluded from the analyses, bringing the total to 86. Of note, there were no CardioMEMS malfunctions or serious adverse events reported in this study.

Figure 1.

INTELLECT 2-HF (Investigation to Optimize Hemodynamic Management of HeartMate II Left Ventricular Assist Device Patients Using the CardioMEMS Pulmonary Artery Pressure Sensor in Advanced Heart Failure) subject follow-up. Disposition of subjects in the INTELLECT 2-HF trial from baseline to 6 months including follow-up at 1 and 3 months as well as reasons for withdrawal. CMEMS indicates CardioMEMS; HMII, HeartMate II; HM3, HeartMate 3; and n, number of subjects.

The enrolled study population of 101 patients had an average age of 61.2±12.5 years, was predominantly male (75%), and 47% had an ischemic etiology of HF (Table 1). Patients had been supported by an LVAD for an average of 20.9±17.8 months before enrollment, with PAP sensors in place for 16.1±12.5 months for the existing CMEMS group. At enrollment, 69% of patients were in NYHA class III and mean 6-minute walk distance was 248±124 m. Hemodynamics at baseline revealed elevated pulmonary artery pressures (PAD 19.8±6.4 mm Hg, PAM 27.1±7.6 mm Hg), with higher PAD and PAM values for patients with new versus existing CardioMEMS. Table 2 shows baseline characteristics of responders (R) and nonresponders (NR) groups. Of the 38 responders, 15/38 (39%) had existing CardioMEMS and 23/38 (61%) had new CardioMEMS devices. R and NR groups were similar except baseline mean PAD was significantly higher in R compared with NR groups (21.5±6.1 versus 18.0±5.8 mm Hg, P=0.008). Figure 2 shows changes in PAD and 6MWD over time. In the R group, the average PAD significantly decreased from baseline to 6 months (21.5 versus 16.5 mm Hg; P<0.001) while in the NR group average PAD increased (18.0 versus 20.3 mm Hg; P=0.002; Figure 2A). At 6 months, the 7-day average PAD was lower in R compared with NR (16.5 versus 20.3 mm Hg; P=0.025; Figure 2A). Similar statistically significant trends were observed in PAS and PAM parameters between the R and NR groups (Figures S1 and S2). The R group demonstrated improved 6MWD from baseline versus 6-months (266 versus 322 m; P=0.025) while those in the NR groups had no change (227 versus 233 m; P=0.666; Figure 2B). Furthermore, at 6 months, the 6MWD was higher in the R versus NR group (322 versus 233; P=0.008; Figure 2B).

Table 1.

Baseline Characteristics: Existing CardioMEMS Versus Newly Implanted CardioMEMS Baseline Demographics

Table 2.

Baseline Characteristics of Responders and Nonresponders and of Subjects With >50% Follow-Up Time in Target Range: PAD <20 Versus PAD ≥20

Figure 2.

Pressure measurement and functional capacity. Pulmonary artery diastolic pressure (PAD) averages (A) and 6-minute hall walk distances (B) for responders compared with nonresponders (NR) at baseline, 1, 3, and 6 months. *P values calculated using 2-tailed t test. †P values calculated using least squares mean.

No differences in pump speeds were identified between R and NR groups, nor between TTR groups for either the HM II or HM 3 LVADs, nor in the durations that the devices were in place before enrollment (Table 2). For the R versus NR groups, there were 278 and 345 interventions over the 6-month study, with most frequent interventions being adjustments in diuretics (65.8% versus 70.8%), followed by vasodilators (42.1% versus 39.6%), and pump speed (18.4% versus 25.0%), respectively (Table 3).

Table 3.

Interventions for Responders Versus Nonresponders

Table 2 also shows the baseline characteristic of subjects who spent >50% of the 6-month study period within the target range of PAD <20 mm Hg (PAD<20) compared with PAD ≥20 mm Hg (PAD≥20). As expected, significantly lower PAD, PAS, and PAM were noted in PAD<20 group compared with PAD≥20 throughout the course of the study (Table 2; Figure 3; Figures S3 and S4). At 6 months, the PAD was lower in PAD<20 group compared with PAD≥20 group (15.6 versus 23.3 mm Hg; P<0.001; Figure 3A). The number of pump speed changes (0.4 versus 1.5 EPPY; P<0.001) and diuretic dose interventions (9.0 versus 15.2 EPPY; P<0.001) were less in the PAD<20 compared with PAD≥20 group (Table 4). Figure S5 shows the changes per-patient-month with up and downtitration of medications and pump speed.

Figure 3.

Hemodynamic measurements, heart failure hospitalization, and urgent care visits. Pulmonary artery diastolic pressure (PAD) averages at baseline, 1, 3, and 6 months for subjects that spent >50% of the study with a PAD less than 20 compared with those with a PAD equal to or greater than 20 mm Hg (A); comparison of all heart failure hospitalizations (B) and urgent clinical visits for volume/cardiovascular medication management (C) for subjects with >50% of time in the target range PAD less than 20 vs PAD greater than or equal to 20 mm Hg. *P values calculated using 2-tailed t test. †P values calculated using least squares mean. <20 indicates PAD less than 20 mm Hg; and ≥20, PAD greater than or equal to 20 mm Hg.

Table 4.

Interventions for Subjects With >50% of Time in the Target Range PAD <20 Versus PAD Greater Than or Equal to 20 mm Hg

The rate of HFH through 6 months was significantly lower in the PAD<20 Group compared with (PAD≥20; 12.0% versus 38.9%, P=0.005; Figure 3B). In addition, there were significantly fewer patients with urgent clinic visits for volume management in PAD<20 versus PAD≥20 (0.0% versus 16.7%, P=0.004; Figure 3C).

Discussion

We conducted the largest, prospective observational study of patients with continuous flow LVADs and implantable hemodynamic PAP monitoring devices. There are several important observations highlighted by the present multicenter clinical investigation. First, the CardioMEMS HF System is compatible with both HeartMate LVADs and can be implanted and reliably used to evaluate pulmonary pressures. Second, there is a high degree of patient compliance with transmission of pressure data, with greater than 90% of patients making at least weekly measurements. Third, relatively modest reductions in PAD (3–5 mm Hg) are associated with significant improvements in 6MWD within the group considered responders and when responders are compared with nonresponders at 6 months. Finally, among patients whose PAD was maintained less than 20 mm Hg compared with greater than or equal 20 mm Hg, significantly fewer HFHs were observed (13% versus 39%; P=0.005). These observations support the conclusion that clinically important HF end points (6MWD and HFH) in LVAD patients are associated to PAP changes (Figure 4).

Figure 4.

Summary figure. One hundred one subjects were enrolled in the INTELLECT 2-HF study (Investigation to Optimize Hemodynamic Management of HeartMate II Left Ventricular Assist Device Patients Using the CardioMEMS Pulmonary Artery Pressure Sensor in Advanced Heart Failure) with both a HeartMate left ventricular assist device (LVAD) and CardioMEMS pulmonary artery pressure sensor and were followed for 6 months. In the patients stratified as responders (R), the average PAD significantly decreased from baseline to 6 months (21.5 vs 16.5 mm Hg; P<0.001) and patients demonstrated an improved 6-minute walk distance at 6 months compared with nonresponders (NRs; who had <1 mm Hg reduction, or increase, in PAD over the 6-month study). In addition, subjects who spent >50% of the study with a PAD less than 20 mm Hg had significantly fewer heart failure hospitalizations compared with those with a PAD equal to or greater than 20 mm Hg throughout the study. *P values by 2-tailed t test, †P values and differences calculated by least squares mean, ‡HFH P value calculated by Poisson regression. <20 indicates PAD less than 20 mm Hg; ≥20, PAD greater than or equal to 20 mm Hg; HFH, heart failure hospitalization; N, number of subjects; and PAD, pulmonary artery diastolic pressure.

While LVAD therapy improves survival and quality of life in patients with advanced HF, 1 in 5 patients do not experience improvement in functional capacity and 1 in 15 patients have recurrent HFHs.^1,2,4–7 Previous investigations including invasive and noninvasive hemodynamic ramp studies have pointed to persistent hemodynamic derangements as a potential target for LVAD speed adjustment or HF medication titration.^6,7,11,22 Additional studies have demonstrated that hemodynamic optimization using ramp protocols is associated with reduced hospitalization and improved functional capacity.^{10,12,23–25} Hemodynamic ramp studies are invasive and time-consuming, however, and both invasive and noninvasive ramp studies reflect resting conditions at a single time point. By contrast, the CardioMEMS system can provide a longitudinal hemodynamic profile of patients supported with LVADs.^26,27 The CardioMEMS pulmonary artery pressure monitoring system has been extensively studied in patients with chronic HF and reduces HFH by 33% to 62%.^13,14,28,29

By design, the INTELLECT-2 study included LVAD patients with persistent HF symptoms that were candidates for medical management optimization guided by PAP measurements. Patients enrolled in this study had been on HM II or HM3 support for an average of 20.9 months. In comparison to the MOMENTUM 3 trial, patients enrolled in the INTELLECT 2-HF trial are more functionally limited after similar duration of LVAD support.⁴ After a comparable duration of LVAD support, 69% of INTELLECT 2-HF patients were NYHA class III at enrollment whereas 80% of the MOMENTUM-3 cohort were NYHA class I or II. Furthermore, the baseline 6-minute walk distance of 248 m in this study was lower than the 323 to 361 m reported for LVAD patients in the MOMENTUM 3 trial after 24 months of support.⁴

Case reports and small series have described the use of CardioMEMS among patients supported with LVADs.^15,18 In these earlier reports, patients were implanted with CardioMEMS before LVAD, similar to the existing CardioMEMS group in our study. The larger of these studies (n=436, 108 patients received an LVAD, and 328 patients did not receive an LVAD) by Kilic et al reported dramatic reduction in PAP after LVAD implantation; however, this study did not prescribe a target PAP or correlate PAP measurements with functional capacity. Veenis et al described 10 patients who underwent CardioMEMS implant immediately before LVAD implantation (HEMO-VAD study).¹⁷ They found that a PAM >25 mm Hg at the time of CardioMEMS implant was associated with an increase in the combined end point of all-cause mortality, renal replacement therapy and right ventricular failure within 1 week after LVAD; however, they did not report on longer-term trends in clinical events or functional capacity.

In our study, a cohort of LVAD patients with NYHA class III symptoms were implanted with CardioMEMS, as per the approved indication (new CardioMEMS group). In this cohort, there were no serious adverse events or safety concerns including bleeding, infection, unsuccessful deployment, or inability to transmit pulmonary pressures. In the overall cohort of 101 patients, we observed a high rate of weekly transmission by patients. Pulmonary pressure waveforms and measurements were not affected by LVAD function.

A total of 42% of patients had CardioMEMS implanted before enrollment in this study with the average duration of 16 months of monitoring. These patients had lower baseline mean and diastolic pulmonary artery pressures as compared with those patients implanted with CardioMEMS after enrollment. The lower baseline pressures may be due to PAP-guided HF management before enrollment, which could limit the possibility of further improvement to qualify as a responder. On the other hand, patients in the new CardioMEMs device group had to meet NYHA Class III indications for use and were therefore more likely to manifest persistent elevation in PAP.

While multiple cardiac and noncardiac mechanisms could result in functional limitation in LVAD patients,³⁰ key hemodynamic determinants include pulmonary and systemic vascular load, left ventricular filling, and pump speed.^7,11,22,23 Elevations in PAP have been implicated in HF disease progression and impending decompensation among patients with HF.^15,16,31,32 Among LVAD patients undergoing hemodynamic ramp studies, the inability to reduce intracardiac pressures is associated with worse clinical outcomes.^9,12 Furthermore, acute improvements in pulmonary pressure during ramp studies is associated with improvements in exercise parameters.^7,23 In our study, we observed that higher PAD (>20 mm Hg) was associated with a higher rate of HFH at 6 months. We also found that modest reductions in PAD (3–5 mm Hg) were associated with improved 6MWD.

Limitations

The current study did not prescribe a treatment algorithm to optimize PAP, and it is noteworthy that speed changes were relatively infrequent. Future study designs could evaluate whether prescribed speed changes and medication adjustments and an invasive baseline hemodynamic ramp study increase the percentage of responders and associated favorable outcomes. The extent to which LVAD speed changes influence pulmonary pressures compared with adjustment of HF medications warrants further investigation. An additional limitation is related to the fact that CardioMEMS does not provide direct information regarding the right heart function or cardiac output. This is particularly relevant in LVAD patients who have high rates of right ventricular failure. Furthermore, enrollment occurred on average 21 months after LVAD implant, which could introduce survivor bias and impact event rates, making it difficult to extrapolate the use of PAP sensors earlier after LVAD implant. Finally, it is conceivable that unmeasured characteristics, such as prior medical treatment or coexisting conditions, explain baseline and follow-up hemodynamic differences between groups.

Conclusions

In conclusion, ambulatory hemodynamic monitoring with the CardioMEMS-HF system in HeartMate LVAD patients appears to be safe, feasible, and reliable. A 3 to 5 mm Hg reduction in PAD at 6 months was associated with significant improvements in 6MWD. Achieving a sustained PAD <20 mm Hg was associated with lower HFH events (Figure 4). These findings provide justification and guidance for a prospective study utilizing the CardioMEMS system to improve the management and outcomes of LVAD patients.

Article Information

Sources of Funding

The INTELLECT 2-HF study (Investigation to Optimize Hemodynamic Management of HeartMate II Left Ventricular Assist Device Patients Using the CardioMEMS Pulmonary Artery Pressure Sensor in Advanced Heart Failure) was funded by Abbott.

Disclosures

Dr Thohan is a consultant for Abbott, Impulse Dynamics, and Battelle and participated in Speaker’s Bureau for Pfizer and Boehringer Ingelheim. Dr Abraham is a consultant for Abbott and Abiomed. A. Burdorf participated in Speaker’s Bureau for Abbott. Dr Pagani works for Scientific Advisory Board for FineHeart Inc. Dr Vidula received research grant funding from Abbott and is supported by National Institutes of Health (NIH) grant 1R01HL155201-01. Dr Heywood received Speaker’s Honoraria and research support from Abbott. Dr Cogswell received Speaker’s Honoraria for Abbott. He is third party employed at Medtronic (spouse). N. Dirckx and Dr Farrar are Abbott Employees. Dr Drakos is a consultant for Abbott. He received research support from Novartis, Merck, NIH, American Heart Association, Department of Veterans Affairs, and Nora Eccles Treadwell Foundation. The other authors report no conflicts.

Supplemental Material

Figures S1–S5

Appendix

List of INTELLECT 2-HF Investigators and Sites: Adam Burdorf, DO, The Nebraska Medical Center, Omaha, NE; Ali Nsair, MD: Ronald Reagan UCLA Medical Center, Los Angeles, CA; Andrew Kolodziej, MD: University of Kentucky, Lexington, KY; Arvind Bhimaraj, MD: The Methodist Hospital, Houston, TX; Brent Lampert, DO, Ohio State University, Columbus, OH; Brian Jaski, MD: San Diego Cardiac Center, San Diego, CA; Christopher Chien, MD: University of North Carolina at Chapel Hill, Raleigh, NC; David Majure, MD: North Shore University Hospital, New York, NY; Francis Pagani, MD: University of Michigan, Ann Arbor, MI; Himabindu Vidula, MD: University of Rochester Medical Center, Rochester, NY; Jacob Abraham, Providence St. Vincent Medical Center, Portland, OR; Jared Herr, MD: California Pacific Medical Center - Van Ness Campus, San Francisco, CA; Joseph Rahman, MD: USC University Hospital, Los Angeles, CA; Melana Yuzefpolskaya, MD: New York-Presbyterian/Columbia University Medical Center, New York, NY; Michael Zacharias, MD: University Hospitals Cleveland Medical Center, Cleveland, OH; Miriam Jacob, MD: The Cleveland Clinic Foundation, Cleveland, OH; Nasir Sulamanjee, MD: Aurora Medical Group, Milwaukee, WI; Noah Moss, MD: Mount Sinai Hospital, New York, NY; Priyesh Patel, MD: WakeMed Hospital, Raleigh, NC; Rebecca Cogswell, MD: University of Minnesota Medical Center Fairview, Minneapolis, MN; Rebecca Napier, MD: Prisma Health-Midlands, Columbia, SC; Rita Jermyn, MD: St. Francis Hospital, Roslyn, NY; Stavros Drakos, MD: University of Utah Health and School of Medicine, Salt Lake City, UT; Thomas Heywood, MD: Scripps Health, La Jolla, CA