Effectiveness of LVAD vs Medical Management in Ambulatory HF
Effectiveness of LVAD vs Medical Management in Ambulatory HF
The success of LVAD therapy, predominantly in patients with INTERMACS profiles 1 to 3, has been well documented in clinical trials. The ROADMAP study, as the first prospective controlled study in patients with advanced ambulatory, noninotrope-dependent HF, found that LVAD therapy significantly improved survival, with improvement in 6MWD ≥75 m at 12 months relative to OMM. The study's observational nature resulted in an imbalance in the severity of illness between the OMM and LVAD patients, as many centers are more likely to delay use of LVAD in the least ill patients. Comparisons of the study arms are made with the acknowledgement that these 2 cohorts had different underlying disease severity. Survival was similar in both groups in the intention-to-treat analysis. As-treated event-free actuarial survival over a 12-month period was significantly better with LVAD than OMM. The LVAD survival rate of 80 ± 4% at 1 year is similar to the recently published post-approval DT HMII study that reported a 12-month survival of 82 ± 5% in patients with INTERMACS profiles 4 to 7 compared with 71 ± 3% for those with INTERMACS profiles 1 to 3. Differences in the primary endpoint between LVAD and OMM were primarily due to the use of delayed LVADs in the OMM group. Thus, the intention-to-treat analysis suggests no mortality penalty for either early or delayed use of LVADs in these patients, indicating that factors beyond survival are paramount to decision-making surrounding LVAD implantation in this population. Patients in the OMM group avoid complex LVAD surgery and LVAD AEs; however, watchful waiting on OMM would not achieve the primary benefits of functional improvements and patient-reported HRQoL with LVAD support. A full examination to understand the benefits and risks of delayed LVAD decision is planned at the 2-year study follow-up.
Pilot data from the medical arm of INTERMACS (MEDAMACS), launched in January 2013, showed that the 1-year event-free survival of death, LVAD, or heart transplant while receiving OMM was approximately 47%. These observations demonstrate that advanced, ambulatory, noninotrope-dependent HF patients with high-grade symptoms of shortness of breath (NYHA functional class IIIB/IV) have a poor 12-month prognosis. Considering that many medically managed HF patients with INTERMACS profiles 4 to 7 could potentially benefit from LVAD therapy, early referral may be warranted to carefully weigh the projected benefits versus risks for individual patients when considering LVAD or continued OMM in this less sick patient population with advanced HF.
Concerns persist that LVADs predispose patients to an undue burden of AEs, including thromboembolic and bleeding events. In the ROADMAP patient population, bleeding, which accounted for two-thirds of all events, was the primary driver of the high event rate in LVAD patients and were almost 9 times the rate of stroke-related thromboembolic events and pump thrombus, which were comparable to findings in past HMII trials. The rates of ischemic and hemorrhagic stroke among LVAD patients (0.06 and 0.03 EPPY, respectively) are similar to those among patients with advanced HF who underwent implantation in the HMII DT trial (0.05 and 0.03 EPPY, respectively) but greater than the rate in patients with advanced HF enrolled in the OMM group of ROADMAP (0.01 EPPY each for ischemic and hemorrhagic stroke). The stroke incidence in those treated with OMM in ROADMAP was lower than that among patients with advanced HF who do not have device support and have other cardiovascular conditions, such as atrial fibrillation.
In the recent past, many centers reduced the targeted international normalized ratio to 2.0 ± 0.5 (quartile 1 to quartile 3: 1.5 to 2.5) for HMII continuous-flow LVADs. The median international normalized ratio in ROADMAP of 2.1 (quartile 1 to quartile 3: 1.8 to 2.5) is higher than that reported in the DT trial of 1.8 (quartile 1 to quartile 3:1.4 to 2.3). In ROADMAP patients, pump thrombosis was only 6.4% at 12 months, which was much lower than in a recent 3-center report but comparable to an INTERMACS investigation. In addition, LVAD patients had a low device-related infection rate (0.14 EPPY), nearly 50% of HMII DT trial patients, but similar to the low driveline infection rate (0.11 EPPY) recently reported in the SSI (Silicone-Skin-Interface) Registry, where the driveline velour was buried below the skin. AEs in the OMM patients were less than one-half that of LVAD patients, and worsening HF (the leading cause) accounted for >80% of events. Not all AEs in the OMM group resulted in rehospitalization, as worsening HF may be treated as an unscheduled outpatient visit in which intensified diuretic or other pharmacotherapy is applied. In contrast, rehospitalizations for the LVAD group are often due to reasons less amenable to outpatient solutions. Many rehospitalizations may be less likely associated with a serious adverse outcome.
Enhancing quality of life and functional capabilities remains a critical therapeutic goal in treating patients with advanced HF. In our study, the use of LVAD compared with OMM was associated with significant reduction in worsening HF-related readmissions, as well as improved quality of life and functional capacity. The exercise and quality of life benefit with LVAD occurred even with AEs being more frequent. Benefits included a 75-m increase in 6MWD, a 30-point improvement in EQ-5D VAS comparable to the HMII DT post-approval study, and the 35-point improvement in INTERMACS data at 1-year follow-up, a reduction in depression, and improvement in symptoms corresponding to a reduction by at least 1 NYHA functional class. The 6MWD absolute improvement at 12 months is also greater than the 45-m increase observed in NYHA functional class IV patients treated with cardiac resynchronization therapy. Compared with the LVAD arm, OMM patients had a higher baseline 6MWD, in addition to other baseline parameters, consistent with a less sick patient population, a marginal 6MWD increase, and a smaller improvement in the EQ-5D VAS. At 12 months, 71% of OMM patients had persistent NYHA functional class III/IV symptoms.
The ROADMAP trial is complementary to the recent National Institutes of Health–funded REVIVE-IT (Randomized Evaluation of VAD Intervention before Inotropic Therapy) trial. The intent of REVIVE-IT was to examine LVAD versus OMM in patients with even less advanced HF (NYHA III symptoms/INTERMACS profile 7) than ROADMAP. However, REVIVE-IT was closed due to failed recruitment and the national principal investigators putting the study on hold due to concern regarding clinical equipoise (projected benefits vs. current AEs associated with LVADs). We hope that ROADMAP provides some clarity regarding the projected benefits versus risks in a less sick patient population (INTERMACS profile 4 to 7), and we remain optimistic that randomized trials will ultimately be designed and completed.
ROADMAP was a nonrandomized, controlled observational study of current practice and decisions, and hence there is potential for bias. Selection bias was not unexpected, as seen in the baseline characteristics, which indicated that OMM patients were appropriately less ill than LVAD patients. Thus, the risk/benefit analyses presented may underestimate the benefit and overestimate the risk of LVAD versus OMM. Despite these differences, more LVAD patients still met the primary endpoint. The withdrawn percentage was 3 times greater in the OMM (9%) versus the LVAD (3%) group and may influence the results. The LVAD used in this study was HMII, and the risk/benefit analyses are not generalizable to other mechanical circulatory support devices. There is well-known bias regarding patient-reported questionnaires and outcomes, including HRQoL and depression. AEs were reported by treating physicians and not adjudicated by a clinical events committee; we do not believe that this affected the study primary endpoint. In addition, important determinants of the appropriateness of LVADs, including frailty, nonadherence to therapy, and social support, were not collected but may have important implications that could skew the results. The study was also performed in a selected patient population, and applicability to the broader population of noninotropic-dependent patients with HF, including those with less hemodynamic and functional compromise than ROADMAP patients, would be speculative. Many findings and endpoints of this observational study should be considered hypothesis generating and need to be confirmed with other studies, including randomized controlled trials where appropriate.
Discussion
The success of LVAD therapy, predominantly in patients with INTERMACS profiles 1 to 3, has been well documented in clinical trials. The ROADMAP study, as the first prospective controlled study in patients with advanced ambulatory, noninotrope-dependent HF, found that LVAD therapy significantly improved survival, with improvement in 6MWD ≥75 m at 12 months relative to OMM. The study's observational nature resulted in an imbalance in the severity of illness between the OMM and LVAD patients, as many centers are more likely to delay use of LVAD in the least ill patients. Comparisons of the study arms are made with the acknowledgement that these 2 cohorts had different underlying disease severity. Survival was similar in both groups in the intention-to-treat analysis. As-treated event-free actuarial survival over a 12-month period was significantly better with LVAD than OMM. The LVAD survival rate of 80 ± 4% at 1 year is similar to the recently published post-approval DT HMII study that reported a 12-month survival of 82 ± 5% in patients with INTERMACS profiles 4 to 7 compared with 71 ± 3% for those with INTERMACS profiles 1 to 3. Differences in the primary endpoint between LVAD and OMM were primarily due to the use of delayed LVADs in the OMM group. Thus, the intention-to-treat analysis suggests no mortality penalty for either early or delayed use of LVADs in these patients, indicating that factors beyond survival are paramount to decision-making surrounding LVAD implantation in this population. Patients in the OMM group avoid complex LVAD surgery and LVAD AEs; however, watchful waiting on OMM would not achieve the primary benefits of functional improvements and patient-reported HRQoL with LVAD support. A full examination to understand the benefits and risks of delayed LVAD decision is planned at the 2-year study follow-up.
Pilot data from the medical arm of INTERMACS (MEDAMACS), launched in January 2013, showed that the 1-year event-free survival of death, LVAD, or heart transplant while receiving OMM was approximately 47%. These observations demonstrate that advanced, ambulatory, noninotrope-dependent HF patients with high-grade symptoms of shortness of breath (NYHA functional class IIIB/IV) have a poor 12-month prognosis. Considering that many medically managed HF patients with INTERMACS profiles 4 to 7 could potentially benefit from LVAD therapy, early referral may be warranted to carefully weigh the projected benefits versus risks for individual patients when considering LVAD or continued OMM in this less sick patient population with advanced HF.
Concerns persist that LVADs predispose patients to an undue burden of AEs, including thromboembolic and bleeding events. In the ROADMAP patient population, bleeding, which accounted for two-thirds of all events, was the primary driver of the high event rate in LVAD patients and were almost 9 times the rate of stroke-related thromboembolic events and pump thrombus, which were comparable to findings in past HMII trials. The rates of ischemic and hemorrhagic stroke among LVAD patients (0.06 and 0.03 EPPY, respectively) are similar to those among patients with advanced HF who underwent implantation in the HMII DT trial (0.05 and 0.03 EPPY, respectively) but greater than the rate in patients with advanced HF enrolled in the OMM group of ROADMAP (0.01 EPPY each for ischemic and hemorrhagic stroke). The stroke incidence in those treated with OMM in ROADMAP was lower than that among patients with advanced HF who do not have device support and have other cardiovascular conditions, such as atrial fibrillation.
In the recent past, many centers reduced the targeted international normalized ratio to 2.0 ± 0.5 (quartile 1 to quartile 3: 1.5 to 2.5) for HMII continuous-flow LVADs. The median international normalized ratio in ROADMAP of 2.1 (quartile 1 to quartile 3: 1.8 to 2.5) is higher than that reported in the DT trial of 1.8 (quartile 1 to quartile 3:1.4 to 2.3). In ROADMAP patients, pump thrombosis was only 6.4% at 12 months, which was much lower than in a recent 3-center report but comparable to an INTERMACS investigation. In addition, LVAD patients had a low device-related infection rate (0.14 EPPY), nearly 50% of HMII DT trial patients, but similar to the low driveline infection rate (0.11 EPPY) recently reported in the SSI (Silicone-Skin-Interface) Registry, where the driveline velour was buried below the skin. AEs in the OMM patients were less than one-half that of LVAD patients, and worsening HF (the leading cause) accounted for >80% of events. Not all AEs in the OMM group resulted in rehospitalization, as worsening HF may be treated as an unscheduled outpatient visit in which intensified diuretic or other pharmacotherapy is applied. In contrast, rehospitalizations for the LVAD group are often due to reasons less amenable to outpatient solutions. Many rehospitalizations may be less likely associated with a serious adverse outcome.
Enhancing quality of life and functional capabilities remains a critical therapeutic goal in treating patients with advanced HF. In our study, the use of LVAD compared with OMM was associated with significant reduction in worsening HF-related readmissions, as well as improved quality of life and functional capacity. The exercise and quality of life benefit with LVAD occurred even with AEs being more frequent. Benefits included a 75-m increase in 6MWD, a 30-point improvement in EQ-5D VAS comparable to the HMII DT post-approval study, and the 35-point improvement in INTERMACS data at 1-year follow-up, a reduction in depression, and improvement in symptoms corresponding to a reduction by at least 1 NYHA functional class. The 6MWD absolute improvement at 12 months is also greater than the 45-m increase observed in NYHA functional class IV patients treated with cardiac resynchronization therapy. Compared with the LVAD arm, OMM patients had a higher baseline 6MWD, in addition to other baseline parameters, consistent with a less sick patient population, a marginal 6MWD increase, and a smaller improvement in the EQ-5D VAS. At 12 months, 71% of OMM patients had persistent NYHA functional class III/IV symptoms.
The ROADMAP trial is complementary to the recent National Institutes of Health–funded REVIVE-IT (Randomized Evaluation of VAD Intervention before Inotropic Therapy) trial. The intent of REVIVE-IT was to examine LVAD versus OMM in patients with even less advanced HF (NYHA III symptoms/INTERMACS profile 7) than ROADMAP. However, REVIVE-IT was closed due to failed recruitment and the national principal investigators putting the study on hold due to concern regarding clinical equipoise (projected benefits vs. current AEs associated with LVADs). We hope that ROADMAP provides some clarity regarding the projected benefits versus risks in a less sick patient population (INTERMACS profile 4 to 7), and we remain optimistic that randomized trials will ultimately be designed and completed.
Study Limitations
ROADMAP was a nonrandomized, controlled observational study of current practice and decisions, and hence there is potential for bias. Selection bias was not unexpected, as seen in the baseline characteristics, which indicated that OMM patients were appropriately less ill than LVAD patients. Thus, the risk/benefit analyses presented may underestimate the benefit and overestimate the risk of LVAD versus OMM. Despite these differences, more LVAD patients still met the primary endpoint. The withdrawn percentage was 3 times greater in the OMM (9%) versus the LVAD (3%) group and may influence the results. The LVAD used in this study was HMII, and the risk/benefit analyses are not generalizable to other mechanical circulatory support devices. There is well-known bias regarding patient-reported questionnaires and outcomes, including HRQoL and depression. AEs were reported by treating physicians and not adjudicated by a clinical events committee; we do not believe that this affected the study primary endpoint. In addition, important determinants of the appropriateness of LVADs, including frailty, nonadherence to therapy, and social support, were not collected but may have important implications that could skew the results. The study was also performed in a selected patient population, and applicability to the broader population of noninotropic-dependent patients with HF, including those with less hemodynamic and functional compromise than ROADMAP patients, would be speculative. Many findings and endpoints of this observational study should be considered hypothesis generating and need to be confirmed with other studies, including randomized controlled trials where appropriate.
Source...