CardioMEMS^® and Remote Hemodynamic Monitoring in Pulmonary Hypertension

Pulmonary hypertension (PH) is a result of precapillary remodeling (isolated precapillary PH) or postcapillary pulmonary venous congestion (isolated postcapillary PH [IpcPH]). Some patients demonstrate combined precapillary and postcapillary PH (CpcPH). Determining the etiology of PH is essential to guide management, and diagnosis is confirmed with invasive hemodynamic assessment with a right heart catheterization (RHC).

Pulmonary arterial hypertension (PAH) is a chronic progressive disease characterized by precapillary PH resulting from adverse remodeling of the pulmonary vasculature, ultimately leading to right ventricular failure. The most recent guidelines from the European Society of Cardiology/­European Respiratory Society define PAH based on mean pulmonary artery (PA) pressure >20 mmHg and pulmonary vascular resistance of >2 Woods units, assuming normal left atrial or wedge pressure.¹  Several risk-stratification tools used in the management of PH, such as the REVEAL 2.0 risk score or the European Society of Cardiology/European Respiratory Society risk stratification table, use hemodynamic variables such as right atrial pressure, pulmonary vascular resistance, cardiac index, and pulmonary arterial ­oxygen saturation, typically obtained by invasive assessment.^1,2  Recognition and management of worsening hemodynamics may allow for earlier intervention and alter prognosis using these risk stratification tools, but repeated invasive assessment carries minor but nonnegligible risks.³ 

In patients with PH, a sophisticated hemodynamic assessment is often required to monitor response to treatment and to prevent the progression of the disease to worsening heart failure (HF). Decongestion strategies in these patients aim at preventing HF hospitalizations, which can ultimately affect mortality.⁴  Traditional nurse-led teams have been very successful in the management of HF.⁵  Wearable and implantable hemodynamic monitors (IHMs) have been developed to determine left atrial pressure, thoracic impedance, and pulmonary pressures.^6–12  Several monitors attached to heart rhythm devices have also been developed and are used commercially but have only had limited success.^13–15  In this article, we discuss the role of the CardioMEMS IHM in patients with left-sided HF including those with PH and in patients with PAH. We discuss the advantages, disadvantages, and future directions in the utility of remote hemodynamic monitoring in PH patients.

CardioMEMS is the only commercially available IHM. It has been studied in several large, randomized trials in left-sided HF, irrespective of ejection fraction (EF), and small pilot studies in patients with PAH. The CardioMEMS system uses microelectromechanical technology with a piezoelectrical membrane to measure PA pressures.¹⁶  Distortion of the piezoelectrical membrane in the sensor changes the resonance frequency signal, corresponding to a pressure shift. This change in pressure shift can be measured with the help of an external measurement system that helps capture data from the implanted device.¹⁶  The frequency at which these measurements will be recorded is at the discretion of the managing provider. The implanted device measures about 45 mm in length and 10 mm in width (Figure 1).

The CardioMEMS IHM was studied first in the CHAMPION trial (2011).¹⁸  The study assessed the rate of HF-­related hospitalizations in patients with New York Heart Association (NYHA) class III symptoms from left heart disease. Both the study and the control cohorts underwent device implantation, but the IHM data were only used for clinical decision making in the study arm.¹⁸  About 78% of the patients had a left ventricular EF (LVEF) of less than 40%, with a mean age of 61 years. At 6 months, a 28% reduction in HF-­related hospitalization in the study group was found (hazard ratio [HR] = 0.70, 95% confidence interval [CI] = 0.60–0.84; P < 0.0001).¹⁸  By the end of the study, at 15 months, the study group had a 39% reduction in HF-related hospitalizations. Freedom from device-related complications was 98.6%, and freedom from sensor failure was 100%, further demonstrating the safety of the device.¹⁸  An open-label long-term follow-up of this population showed a persistent reduction in HF-­related hospitalizations at an additional 13 months.¹⁹  Compared with the prior randomized duration, this control group had a significant reduction in HF-­related hospitalizations when their hemodynamic data were available to their providers (HR = 0.52, 95% CI = 0.40–0.69).¹⁹  This study led to U.S. Food and Drug Administration (FDA) approval of the CardioMEMS system in 2014 for patients with NYHA class III symptoms with one HF-related admission in the year preceding ­implantation.²⁰ 

A retrospective analysis of the CHAMPION trial population was performed to assess the outcomes concerning World Health Organization (WHO) group II PH.²¹  Patients without PH were at lower risk for mortality (HR = 0.31, 95% CI = 0.19–0.52, P < 0.0001) and lower risk of hospitalization (0.37/year versus 0.77/year, HR = 0.49, 95% CI = 0.39–0.61, P < 0.001).²¹  A significant reduction in HF hospitalizations in patients with and without PH was found; however, in patients with PH, a reduction in the composite endpoint of death and HF hospitalizations was found with access to IHM data (HR = 0.74, 95% CI = 0.55–0.99, P = 0.04) but no difference in survival (HR = 0.78, 95% CI = 0.50–1.22, P = 0.28).²¹  Further, in the CHAMPION trial population, a 30% increase in mortality for every 5 mmHg increase was found in PA pressure in patients with an EF of about 68%.¹⁹ 

The GUIDE-HF trial was a second randomized trial assessing the benefit of CardioMEMS.²²  Inclusion criteria in this trial included patients with NYHA class II–IV symptoms with a recent HF hospitalization and/or elevated natriuretic peptides.²²  Like the CHAMPION trial, the device was implanted in the study and control groups. The PA pressure measurements of the control group were unavailable for review in clinical decision making. The primary endpoint consisted of all-cause mortality and total HF events, including HF hospitalizations and urgent HF visits.²²  The study failed to meet superiority for its primary endpoint (HR = 0.88, 95% CI = 0.74–1.05). Also, no significant reduction in prespecified HF-related events was found (HR = 0.85, 95% CI = 0.70–1.03). The study enrollment and follow-up were affected by the COVID-19 pandemic.²²  A pre-COVID-19 sensitivity analysis, however, showed a reduction in HF event rate (HR = 0.76, 95% CI = 0.61–0.95).²² 

The MEMS-HF study was a prospective, nonrandomized study that enrolled patients with NYHA class III symptoms who had an HF-related hospitalization the preceding year.²³  During the first 6 months after implantation, HF hospitalizations decreased by 62% (HR = 0.38, 95% CI = 0.31–0.48), and over the 12 months of follow-up, HF hospitalizations decreased by 66% (HR = 0.34, 95% CI = 0.26–0.44).²³  Patient-reported quality of life scores, including the Kansas City Cardiomyopathy Questionnaire (KCCQ), EQ-5D-5 L questionnaire, and the Patient Health Questionnaire depression module, were assessed and showed improvement at 6 months that persisted at 12 months.²³ 

A subanalysis of the MEMS-HF study was performed, assessing for any difference in outcomes based on the presence of PH.²⁴  A total of 106 patients’ RHC tracings were analyzed, and they were divided into 3 groups: no PH (31 patients), IpcPH (38 patients), and CpcPH (36 patients).²⁴  During the 12-month follow-up, the systolic, diastolic, and mean PA pressures decreased in the latter 2 groups, whereas the mean and ­diastolic PA pressures decreased in patients without PH.²⁴  HF hospitalization reductions were comparable in the CpcPH group (0.639 events/patient-year; HR = 0.37) and the IpcPH group (0.72 events/patient-year; HR = 0.45).²⁴  The group without PH had the most significant benefit (0.26 events/patient-year; HR = 0.17, P = 0.04 versus IpcPH/CpcPH groups).²⁴  A substantial improvement in quality of life and NYHA class was found in all the subgroups.²⁴ 

Remote hemodynamic monitoring of PA pressures has also been performed in patients with left-sided HF to assess the effect of empagliflozin in the EMBRACE-HF trial. This study showed that empagliflozin significantly reduced PA diastolic pressures as assessed with the CardioMEMS system, and this effect was seen as early as 1 week after drug initiation.²⁵ 

The MONITOR-HF study assessed for quality-of-life improvement and reduced HF-related hospitalizations in patients with remote hemodynamic monitoring with CardioMEMS versus standard of care in Europe.²⁶  This study enrolled participants with NYHA class III symptoms irrespective of baseline EF.²⁶  It was a randomized, open-label multicenter trial. The trial enrolled 348 patients with a median age of 69 years.²⁶  Fifty percent of the patients had ischemic cardiomyopathy, and 27.9% had LVEF greater than 40%.²⁶  The primary outcome was a mean change in KCCQ score at 12 months. At 12 months, mean changes of +7.05 (95% CI = 2.77–11.33) for the study group and −0.08 (95% CI = −3.76 to 3.60) with P = 0.013 for the control group were found in KCCQ. No significant difference in cardiovascular death or all-cause mortality was found. However, the NT-proBNP and 6-minute walk distances at 12 months significantly differed, favoring the study group.²⁶ 

Table 1 summarizes trials that assessed the safety and clinical efficacy of IHMs.

The INTELLECT 2-HF study is a multicenter, prospective, nonrandomized, observational study that assessed the feasibility and clinical utility of CardioMEMS in patients with durable left ventricular assist devices (LVADs).²⁷  Fifty-two patients with HeartMate II and 29 patients with HeartMate 3 LVADs with existing or newly implanted CardioMEMS sensors were followed for a total of 6 months.²⁷  The population was stratified into responders (average reduction of PA diastolic pressures by at least 1 mmHg over 6 months) and nonresponders (average decrease of PA diastolic pressures by less than 1 mmHg over 6 months). A significant improvement in 6-minute walk distance among responders (266 m versus 322 m; P = 0.025) was found compared with no change in nonresponders. Further, patients whose PA diastolic pressure was less than 20 mmHg for over half of the study duration had a significantly lower rate of HF-related hospitalization (12% versus 38.9%; P = 0.005).²⁷ 

Trials studying the use of CardioMEMS IHMs looked at outcomes irrespective of EF. However, a GUIDE-HF subanalysis assessed the outcomes by EF in guideline-defined subgroups of EF ≤ 40%, 41%–49%, and ≥50%.²⁸  A ­bimodal distribution of LVEF was found, with the majority in the HF with reduced EF (53%) and HF with preserved EF (HFpEF; 40%) subgroups. Patients in the HFpEF subgroup tended to be White, older, female, and with higher body mass index, higher blood pressure, lower rates of coronary artery disease, and lower estimated glomerular filtration rates. The NT-proBNP, KCCQ12 scores, and 6-minute walk distances were similar across the spectrum.²⁸  Across all subgroups, a significant reduction was found in the primary endpoint of composite HF hospitalizations, urgent care visits related to HF, and all-cause mortality. These results seen in the HFpEF population, unlike drug therapy studies in which the efficacy of drugs alone declines in reaching these same primary endpoints, indicate the importance of targeting filling pressures to reduce morbidity and mortality.²⁸ 

While the above-noted studies assessed the safety and utility of CardioMEMS in left-sided HF patients, many of whom had secondary PH, a small proof-of-concept pilot study evaluated the safety and feasibility of implantation in the PAH population.²⁹  In this study, the CardioMEMS device was implanted in 27 patients with PAH and NYHA class III or IV symptoms.²⁹  The device was implanted in the cardiac catheterization lab and calibrated with RHC PA pressures and thermal and/or indirect Fick cardiac output (CO) assessment.²⁹ 

CardioMEMS-derived CO was calculated using a proprietary algorithm based on PA pressure waveform, Pulmonary artery pressures, heart rate, and reference CO. The total pulmonary resistance could also be calculated based on device parameters. The mean age in their study was 51 years; 92% were women, and 81% had NYHA class III symptoms.²⁹  The median weekly transmission compliance was 98.2%.²⁹  Of the 28 attempted implant procedures, 1 procedural complication occurred with microperforation of the PA during the ­predeployment angiogram before any attempt at CardioMEMS deployment, and he or she died from this complication. Of the successful implants, no device-­related serious adverse events occurred.²⁹  Long-term follow-up data showed a reduction in the average number of RHC per year (4.37 prior to device implant versus 1.99 postdevice implants).³⁰  All the devices remained functional. However, 8 patients required repeat RHC for device recalibration.³⁰ 

The ARTISAN study is a prospective, multicenter, open-labeled trial currently enrolling patients to evaluate the effect of early and rapid treprostinil therapy to reduce mean PA pressures. In this trial, the CardioMEMS system is being used to follow the mean PA pressures of the participants. This study is expected to be completed in September 2024.³¹  Another ongoing phase 2 study assessing a novel drug CS1 uses the CardioMEMS system to follow the participants’ PA pressures.³² 

The most recent guidelines for the management of HF were released in 2022 by the American College of Cardiology and the American Heart Association. These guidelines suggest that, in selected adult patients with NYHA class III HF and a history of HF hospitalization in the preceding year or elevated natriuretic peptide levels, while on maximally tolerated guideline-directed medical therapy, the usefulness of IHMs remains uncertain to reduce the risk of subsequent HF hospitalizations (2b, level of evidence B-R).³³  At the time of this review, no recommendations regarding the use of IHMs in patients with PAH exist.

The Chronicle IHM (Medtronic) was studied in the COMPASS-HF trial (2008).¹²  This prospective, multicenter, randomized, single-blind, parallel-controlled trial was conducted among 274 patients with NYHA class III or IV symptoms.¹²  Primary endpoints were freedom from system-related complications, freedom from pressure-sensor failure, and reduction in HF-related events.¹²  The primary efficacy endpoint did not reach statistical significance despite the study group having a 21% lower rate of all-HF-related events.¹²  After the study, the device failed to receive FDA approval for commercial use.

The Cordella PA pressure sensor (Figure 2) has been developed more recently. It was first studied for patients with HF, irrespective of their EF, with NYHA class III symptoms in the SIRONA first-in-human study.³⁴  The study’s primary safety endpoint was freedom from device-related adverse events through 30 days postprocedure, and the primary efficacy endpoint was accuracy of the device PA pressure measurements, compared with a RHC.³⁴  A total of 15 patients underwent device implantation with a mean age of 71.4 years; 67% were male, and 53% had an EF greater than 40%.³⁴  At 90 days, no device-related adverse events occurred. At 90 days postimplantation, the primary efficacy endpoint was met in all patients.³⁴ 

Subsequently, in the SIRONA 2 trial, the accuracy of the PA sensor was assessed compared with RHC.³⁵  In a total of 70 participants, the equivalence between the PA sensor and RHC for mean PA pressures was excellent, with measurements within equivalent bounds of −4 to 4 mmHg (P = 0.003).³⁵  The device safety profile was excellent, with 98.6% freedom from device-related complications and no reported pressure sensor failures.³⁵  Long-term follow-up of the SIRONA 2 cohort showed that, at 12 months, good agreement between the Cordella PA sensor and RHC continued, with the average difference for mean PA pressure being 2.9 ± 7.3 mmHg.³⁶  No pressure sensor failures were found, and 98.4% freedom from device/system-­related complications occurred.³⁶  The device is now being studied for clinical effectiveness in a single-arm, PROACTIVE-HF trial.¹¹ 

An advantage of the Cordella sensor is that it has no leads and does not require batteries. It is implanted in the right PA and interrogated via an external antenna in the handheld reader.³⁴  The specialized anchor design allows the sensor to be placed into an anterior branch, which along with its microelectronic mechanical system, makes reading from the anterior chest wall possible.³⁴  Pressure applied to the sensor causes deflections of the pressure-sensitive surface, resulting in a shift in the resonant frequency, which can be measured.³⁴ 

Expert guidelines recommend periodic hemodynamic monitoring for risk stratification and management of decisions, especially with persistent symptomatology and failure to respond to medical management.^1,33  Patients may require repeated hemodynamic assessment within a relatively short period, which can be cumbersome to tolerate and carries a small but identifiable risk of complications. In addition, remote monitoring can provide additional data regarding patients’ hemodynamics in their home environment, which may be more reflective of their day-to-day hemodynamic burden of PH than an isolated RHC. Patients can also transmit their PA pressure readings from wherever they are, with the thought that a change in pulmonary pressures can be identified before the onset of symptoms and may provide any warning sign of decompensation, especially in patients who may live a distance away from their PH clinicians.

Further, while remote hemodynamic monitors provide only partial hemodynamic data, they offer a reduction in the number of invasive RHCs a patient may need to undergo in their lifetime. These may also increase access to health care among patients in rural communities where access to health care may be limited due to several reasons, as evidenced by an increase in HF hospitalizations among rural communities.³⁷  The rapid growth of telehealth services secondary to the COVID-19 pandemic provides an opportunity for us to address health care equity among these underserved populations,³⁸  and IHMs can play a vital role in achieving this.

In addition to the risks associated with the implantation procedure itself, additional considerations exist. Despite having an IHM, these patients continue to require periodic RHC for complete hemodynamic assessment. Clinical worsening can be missed based on remote monitoring alone. It may be challenging to detect decompensated right HF without a parallel rise in pulmonary pressures in those falling off the Frank-Starling curve. Remote monitors measure pulmonary pressures at 1 time point when the patient is recumbent, so trends may not accurately reflect change with exercise.

The CardioMEMS system has proven to be cost effective, assuming that the trial outcomes are sustained and that the durability of the device stands. Real-world postmarketing surveillance data on durability and long-term outcomes would clarify its value.³⁹  Also, establishing a remote hemodynamic monitoring program requires immense institutional support. The device is implanted either by an interventional cardiologist or an advanced HF cardiologist. Programs then require dedicated coordinators or program managers to follow these patients closely and adjust therapy based on IHM readings with their providers. Given the variability in reimbursement patterns across payors and regions, assessing the number of patients required for programs to breakeven or even turn profitable is challenging.

In addition to IHMs, several novel devices are being studied for use in patients with PH. For instance, the use of accelerometers is being evaluated as an alternative to the traditional 6-minute walk distance as an objective measure of physical activity in patients with PH.⁴⁰  When remotely monitored by these devices, a reduction in physical activity has been hypothesized to precede clinical worsening and can help identify patients at risk of future hospitalizations.⁴⁰  Behavioral change techniques such as text message–based reminders have been shown to improve physical activity levels, ultimately leading to improved quality of life.⁴¹  Using IHMs to complement other devices as a part of a larger digital ecosystem could be additive in improving patient care and quality of life.

Hemodynamic assessment is vital in the diagnostic classification and management of patients with PH. IHMs play a key role in managing patients with left-sided congestive HF and concomitant PH. Early intervention is possible based on timely recognition of decompensation prior to clinical deterioration, as seen in several studies. IHM use in PAH patients is still in its relative infancy. However, device implantation appears to be relatively safe and feasible in most PAH patients and, in theory, can play a pivotal role in rapid medication titration. Ideally, an implantable device that provides comprehensive hemodynamic data will allow for better management of our patients with PH, helping us manage both left- and right-sided HF.