Fig. 15.1
Taken from the REMATCH trial , this Kaplan-Meier analysis depicts a reduction of 48% in the risk of death from any cause (primary endpoint) in the group that received the LVAD versus the group that received optimal medical therapy alone [7]
At the time of its publication, the REMATCH trial was noteworthy for having enrolled heart failure patients with the most severe clinical and hemodynamic compromise and mortality rates to date [1, 9]. An unanticipated 71% of patients were on inotropic infusions at the time of randomization [9]. Post hoc analysis of the outcomes of patients on and off intravenous inotropic therapy at the time of randomization confirmed that the patients on inotropes derived a near doubling of survival benefit from LVAD implantation with a 1-year survival of 49% and 24% in the LVAD and OMT groups, respectively (p = 0.0014) [9]. Survival of patients on baseline inotrope therapy was equal to or better than that in the OMT group at all times, even in spite of the predicted excess of perioperative mortality due to LVAD implantation [9]. The difference in 1-year survival rates was, however, not statistically significant between the LVAD (57%) and OMT (40%) groups who were not on intravenous inotropic support at baseline (p = 0.55) [9]. It was thus concluded that LVAD implantation was most beneficial for the sickest patients with advanced heart failure.
Enhancements in LVAD design based on lessons learned from the REMATCH trial led to the development of the HeartMate XVE (HM XVE) . In a nonrandomized, prospective trial, Lietz et al. studied the outcomes of 280 patients who were implanted with the modified HM XVE between 2001 and 2005 for DT [1]. This study aimed to investigate the impact of this improved pulsatile-flow LVAD on DT outcomes and to identify clinical predictors portending worse prognosis that could then be made into a risk score to stratify DT candidates. Rates of survival were 86.1%, 56.0%, and 30.9% at 30 days, 1 year, and 2 years after LVAD implantation, respectively [1]. The high 1-year survival rates among recipients of HM XVE as DT were corroborated by a smaller, albeit nonrandomized trial. This study by Long et al. compared the survival of 48 recipients of the HM XVE at four of the highest volume centers participating in the Thoratec DT registry with that seen in the historical LVAD arm of the REMATCH trial [10]. It found a statistically non-significant but nonetheless remarkable 40% decline in the rate of death per patient year from any cause in the HM XVE arm as compared to the HM VE arm of the REMATCH trial [10]. The lesser mortality rates were attributed to improved LVAD design and patient management protocols from years of experience.
While pulsatile-flow LVADs had earned acceptance as therapy for refractory heart failure, their limited long-term durability and large pump size created a need for the simpler and smaller design of continuous-flow LVADs (CF-LVADs) [4, 11]. The HeartMate II trial was among the earliest of studies that looked at the outcomes of CF-LVADs, which were being tested only as a bridge-to-transplant at the time. By 180 days of the study, 100 of the 133 patients implanted with a HM II had either undergone cardiac transplantation, had significant cardiac recovery, or were still on ongoing mechanical support while remaining eligible for transplantation [11]. Although the results of a study on BTT patients cannot be directly applied to DT patients, as the degree of severity of heart failure and associated comorbidities making a patient ineligible for transplant and hence a DT patient make for a much sicker population, this study did show the promise of CF-LVADs .
The staggering results of REMATCH and other trials individually comparing the survival benefit among patients implanted with improved pulsatile-flow LVADs and continuous-flow LVADs begged the need for a head-to-head comparison of pulsatile-flow and CF-LVAD therapies. Conducted between 2005 and 2007 and using similar eligibility criteria as the REMATCH trial which have since been adapted into the CMS (Center for Medicare and Medicaid Services ) criteria [12], this multicenter trial randomized 200 DT patients to either CF-LVAD (HeartMate II) or pulsatile-flow HeartMate XVE groups [13]. The primary endpoint of a composite of survival at 2 years, free of disabling stroke (stroke with a Rankin score > 3) or the need for reoperation to replace the device, was achieved by 46% and 11% in the HM II and HM XVE groups, respectively, with p < 0.001 [13]. Subgroup analysis of 1- and 2-year survival rates showed similarly significant results of 68% and 58%, respectively, in the CF-LVAD group and 55% and 24%, respectively, in the pulsatile-flow LVAD group (Fig. 15.2) [13].
Fig. 15.2
Taken from the HeartMate II destination therapy trial , this Kaplan-Meier analysis depicts the striking superior survival rates seen up to 24 months after implantation of the continuous-flow HeartMate II compared to the pulsatile-flow HeartMate XVE [12]
Adverse Events
While both continuous and pulsatile-flow LVADs demonstrated significant survival benefits in patients with end-stage heart failure who were ineligible for transplantation, their use was not without complications. In the REMATCH trial, for instance, patients randomized to the device group were 2.35 times as likely as those in the OMT group to suffer a serious adverse event [8]. Infection, suspected malfunction of the LVAD, non-neurologic bleeding, and neurologic dysfunction were the most common adverse events associated with the LVAD group [8].
Infection
Infection was by far the most frequent adverse event in LVAD studies and, in the era of pulsatile pumps, was particularly commonplace. The REMATCH trial stated that 28% of patients implanted with HM VE developed infection within the first 3 months, with most cases related to a local driveline tract and pocket infection. While most could be treated with local antibiotics, sepsis still claimed 17 of 68 lives [8]. A comparison of the results of HM XVE and HM II recipients showed that CF-LVADs significantly reduced the rates of LVAD-associated infections by 50% as well as that for non-LVAD-associated local infections and sepsis [13]. The larger area of surgical dissection required for implantation of a pulsatile-flow LVAD and the percutaneous driveline were felt to be the most likely causes of increased infection risk associated with pulsatile-flow pumps [11, 13, 14]. Additionally, CF-LVADs lack the compliance chambers, polyurethane membranes, and prosthetic valves that may become niduses for infection [15]. Indeed, the transition to the use of CF-LVADs and the higher cumulative experience of LVAD management at high-volume centers correlated with reduced overall LVAD-associated infections. Institutional changes to use management guidelines, abdominal binders to stabilize the percutaneous driveline, and antibiotic prophylaxis likely contributed to reduced rates of infection with LVADs over the years [10, 14, 16].
LVAD Dysfunction
Suspected LVAD dysfunction was the second most common adverse event seen in the REMATCH trial, amounting to a 35% probability of device failure in 24 months. Within the device group, ultimately, 10 of the 68 patients required replacement [8]. Largely due to improvements in LVAD design, CF-LVADs unsurprisingly fared better in this regard. As the HeartMate II study showed, fewer patients in the HM II group required reoperation to repair or replace the pump than those in the HM XVE group (p < 0.001) [13].
Non-neurologic Bleeding
Non-neurologic bleeding has also been a vexing problem with LVADs. Balancing the risk of device-related thromboembolic events with an increased propensity for bleeding due to anticoagulation and the theorized increased development of acquired von Willebrand syndrome and arteriovenous malformations in the setting of chronically low pulse-pressure has been a struggle with LVAD management [17, 18]. In the REMATCH study, even despite a lack of routine anticoagulation, the frequency of bleeding within the first 6 months following implantation with the HM VE was as high as 42% [8]. During the early trials using HM II for BTT, some centers had adopted a stringent antiplatelet and anticoagulation regimen of aspirin and dipyridamole with a postoperative heparin bridge to warfarin to achieve a target international normalized ratio or INR of 2.0–3.0 [11]. But this regimen resulted in a significantly increased rate of bleeding events, especially in the early postoperative period [11]. A comparison of HM II with HM XVE outcomes showed that both types of LVAD were associated with ten times the rate of bleeding as thromboembolic events, even though only those assigned to the HM II group were anticoagulated [13]. Based on these results, many centers have since reduced the targeted INR to 1.5–2.5 for continuous-flow LVADs and have eliminated the heparin bridge [14].
Neurologic Events
Surprisingly, despite a lack of routine anticoagulation use in the LVAD group of the REMATCH trial, 76% of patients were free of serious neurologic events [8]. Only 10% of patients in the device group suffered an ischemic stroke [8]. The low percentage was attributed to the textured surfaces of the HM VE [8]. Comparison of the HM XVE and HM II recipients showed similar rates of ischemic stroke between the two groups of 14% and 17%, respectively [13]. In fact, while on Coumadin targeted to an INR of 2.0–3.0, the HM II patients suffered a similar rate of ischemic stroke as that among other patients with advanced heart failure and atrial fibrillation who are not on device support [13].
Overall, head-to-head assessment of pulsatile and continuous-flow LVADs showed significant reductions in the rate of major adverse events among CF-LVAD patients including device- and non-device-related infections, RV failure, respiratory failure, renal failure, and cardiac arrhythmia [13]. The study concluded that implantation of CF-LVADs as compared to pulsatile-flow devices among advanced heart failure patients being considered for destination therapy significantly improved the chance of survival free of stroke and the need for reoperation for device repair or replacement at 2 years [13].
An understated confounder of the results which showed an overwhelming superiority of CF-LVADs when compared to pulsatile-flow models was the increased clinical experience that was gained during the early years of device development in the management of these complicated patients. A follow-up study using the same patient pool from the REMATCH trial but followed for an additional period of 375 patient-months found lesser rates of adverse events among the group of patients that had been enrolled during the second half of the trial [14]. A study of 377 patients who were implanted with the HeartMate I generation of LVADs (HM VE and XVE) for DT a rising trend in 1-year survival rates related to the center volume of DT surgeries performed [16]. However, when preoperative DT risk score adjusted analysis of 1-year survival rates showed that DT center volume was not an independent predictor of survival, it was surmised that other factors related to center experience such as better patient selection and improved perioperative treatment accounted for the difference [19].
Causes of Death
The primary causes of death among patients with the pulsatile HeartMate VE were sepsis (41%) followed by LVAD failure (17%) instead of terminal heart failure which accounted for the majority of deaths in the optimal medical therapy group of the REMATCH trial [8]. A trial looking at HM XVE recipients in the post-REMATCH era found that the majority of in-hospital deaths (79%) occurred within the first 3 months [1]. During a minimum follow-up period of 2 years after LVAD implantation, the leading cause of death was hemorrhagic stroke which was seen in 9 and 10% of patients in the pulsatile and continuous-flow recipients, respectively [13]. This was followed by right ventricular failure which occurred in 8 and 5% of the pulsatile and continuous-flow LVAD recipients, respectively [13]. External power interruption, respiratory failure, cardiac arrest, and bleeding each accounted for 3–4% of the rest of the mortalities in the HM II group [13]. The estimated 1- and 2-year survival rates in this group were 68% and 58%, respectively [13].
Risk Score for Survival to Hospital Discharge and 1-Year Post HM XVE
Risk scores and risk calculators have become an important tool in patient selection for LVAD and in the education of patients that are interested in LVAD. In the cases in which the LVAD is used as destination therapy, the patient as well as the physician must be aware of risk for in-hospital mortality and 1-year outcomes. The authors of the post-REMATCH era [1] developed a risk score that can assist the discussion. Although the HeartMate XVE LVAD was used during the trial, the data does support the notion that sicker patient pre-LVAD implant does worse. Using univariate and multivariable analyses, the results were able to predict the risk factors for 90-day in-hospital mortality following LVAD. Table 15.1 shows the nine variables used in the multivariable model.
Table 15.1
Multivariable analysis of risk factors for 90-day in-hospital mortality after LVAD as DT [6]
Patient characteristics | Odds ratio (CI) | p | Weighted risk score |
|---|---|---|---|
Platelet count ≤148 × 103/μL | 7.7 (3.0–19.4) | <0.001 | 7 |
Serum albumin ≤3.3 g/dL | 5.7 (1.7–13.1) | <0.001 | 5 |
International normalization ratio >1.1 | 5.4 (1.4–21.8) | 0.01 | 4 |
Vasodilator therapy | 5.2 (1.9–14.0) | 0.008 | 4 |
Mean pulmonary artery pressures ≤25 mmHg | 4.1 (1.5–11.2) | 0.009 | 3 |
Aspartate aminotransferase >45 U/mL | 2.6 (1.0–6.9) | 0.002 | 2 |
Hematocrit ≤34% | 3.0 (1.1–7.6) | 0.02 | 2 |
Blood urea nitrogen >51 U/dL | 2.9 (1.1–8.0) | 0.03 | 2 |
No intravenous inotropes | 2.9 (1.1–7.7) | 0.03 | 2 |
Each variable was assigned a weighted risk score and the cumulative risk score was calculated for each patient. A maximum score that can be obtained was 31. The patients were then divided into four operative risk categories based on probability of mortality during the hospitalization. Table 15.2 and Fig. 15.3 provide the in-hospital mortality and survival for DT-LVAD patients receiving a HeartMate XVE LVAD .
Table 15.2
Operative risk categories and risk score for 90-day in-hospital mortality after LVAD implantation as DT, survival to hospital discharge, and 1-year survival [6]
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
