Overcoming the resistance to impedance monitoring in heart failure patients

This editorial refers to ‘Device-based impedance measurement is a useful and accurate tool for direct assessment of intrathoracic fluid accumulation in heart failure’ by J. Becher et al ., on page 731 and ‘Intrathoracic impedance and pulmonary wedge pressure for the detection of heart failure deterioration' by M. Maines et al ., on page 680. Patients with moderate to severe heart failure who meet the criteria for implantation of a cardiac resynchronization therapy (CRT) device and/or an implantable defibrillator (ICD) carry a startling risk of suffering from adverse clinical events, in particular heart failure-related hospitalizations and death. Though CRT and ICDs improve survival and CRT also decreases morbidity, the event risk in device carriers remains high: while ICD treatment per se does not impact heart failure severity—from the perspective of the devil's advocate, it may contribute to prolonged suffering from recurring heart failure exacerbation—even many CRT recipients continue to suffer from frequent heart failure-related events, especially in the unfortunate group of non-responders. Therefore, increasing attention has recently been paid to the use of heart failure devices not only for therapeutic but also for diagnostic purposes. Device-based sensors and algorithms for continuous surveillance of various physiological parameters such as the heart rate, heart rate variability, physical activity, or arrhythmia incidence, may add valuable information to patient assessment and, thus, help to improve patient management in heart failure. In particular, diagnostic features of monitoring of cardiac filling pressures and pulmonary fluid retention have raised interest due to their potential to reveal changes in heart failure status timely and early enough to initiate appropriate therapeutic interventions and prevent further progression to overt cardiac decompensation. Both methods showed promising results in pilot studies, 1 , 2 encouraging the design of large-scale clinical trials 3 , 4 and, in the case of impedance monitoring, the incorporation into commercially available ICDs. Monitoring of right ventricular pressures is technically complicated, because the implantation of an additional sensor lead or a separate haemodynamic monitoring system is required. In contrast, intrathoracic impedance can be easily measured between the coil of a right ventricular ICD-lead and the device-can and reflects the electrical conductance across the interpositioned thoracic organs, which improves e.g. with increasing pulmonary fluid retention and thoracic blood volume. The basic principle behind this technology has been studied already decades ago using transthoracic monitoring equipment 5 or minute ventilation sensors for rate adaptive pacing. 6 However, despite the widespread use of ICDs with capabilities of intrathoracic impedance monitoring, there is still considerable resistance to its routine clinical application. This is in part due to concerns about the reliability and diagnostic accuracy of impedance based fluid alert algorithms and the lack of trial evidence to support routine clinical use. Therefore, further investigations into the subject that clarify the pathophysiological correlates of impedance changes and promote the refinement of monitoring technology are clearly warranted. In this volume of the Journal, two articles are published that improve our understanding of intrathoracic impedance monitoring in heart failure. First, Becher et al . 7 present results from an animal study confirming basic principles of the method. The authors nicely document the gradual decrease in the intrathoracic impedance associated with the development of elevated cardiac filling pressures and accumulation of extravascular lung water during cardiac decompensation. For this purpose, acute pulmonary oedema was induced in 12 sheep with previously healthy hearts. By infusing high doses of noradrenalin and a combination of plasma-expander and crystalloid fluids, blood pressure, and cardiac output was increased manifold which resulted in massive haemodynamic overload and retention of extravascular pulmonary fluids. As these experimental conditions apparently differ from those applying to the protracted development of cardiac decompensation in patients with chronic heart failure, conclusions as to the clinical usefulness of impedance monitoring in humans are difficult to draw. An interesting aspect of the study is, however, the comparison of bipolar and tripolar electrode configurations. Although the difference in impedance changes between these configurations was small, the authors describe significant diagnostic benefits with the tripolar method. This underlines the potential of improving impedance-based heart failure diagnostics by further technological development of the measurement equipment and the related software algorithms. Such advancement may also include the use of multiple impedance vectors involving the left ventricular lead in patients with biventricular pacing. 8 Notably, frank pulmonary oedema is easily identified by standard clinical investigation and no device-based feature is necessary to facilitate such obvious diagnosis. In contrast, the domain of implantable heart failure monitoring is the detection of minor and perhaps sub-clinical haemodynamic changes that are easily missed by conventional diagnostic tools. Therefore, the observations during the early phase of lung oedema induction deserve special attention. During the initial stages of progressive decompensation, relative changes in cardiac filling pressures were considerably larger compared with concomitant decreases in the intrathoracic impedance. This might be expected since a certain pressure increase is required to start extravasion of fluids into pulmonary tissues and, eventually, the retention of lung water when the capacity of lymphatic drainage is overwhelmed. These observations open for the view that haemodynamic changes in the early phase of an impending decompensation are better revealed by measurements of filling pressures than impedance. Alternatively, this may indicate that the inclusion of information from central pressure measurements into the fluid detection algorithm may increase the diagnostic value of impedance monitoring. The latter hypothesis is addressed by Maines et al . 9 They evaluated the relationship between non-invasively estimated pulmonary capillary wedge pressure (PCWP) and the intrathoracic impedance in 23 heart failure patients with an ICD. During 23 ± 11 months of follow-up, 45 paired assessments were performed yielding a good agreement between abnormal PCWP estimates (>13 mmHg) and the impedance derived OptiVol fluid index ( k = 0.701, P < 0.001). Furthermore, PCWP and OptiVol values were significantly correlated ( r = 0.677, P < 0.001). These results corroborate previous findings in patients treated with intravenous diuretics in the intensive care setting 2 and demonstrate that the relationship between cardiac filling pressures and the intrathoracic impedance is also present during long-term follow-up in ambulatory patients. Furthermore, PCWP was elevated in all patients presenting with worsening heart failure leading the authors to suggest that a detection algorithm based on the combined analysis of PCWP and impedance information may reduce the number of false alerts and improve the positive predictive value. Yet, owing to the small sample size and event rate, the study does not provide sufficient power to assess the diagnostic performance of the OptiVol fluid detection algorithm. This objective has been addressed in an international multicenter study (SENSE-HF) that is expected to report in the near future. 10 Of note, the present study findings are based on secondary data from the impedance derived OptiVol fluid index, an algorithm that calculates the accumulating difference between an auto-calibrated reference impedance and the actual patient impedance over time. No data on raw-impedance values is reported which would be of greater interest, as it reflects the primary physiological information recorded by the device. However, the agreement between the PCWP and OptiVol values in this study emphasises the basic usefulness of impedance monitoring to track relevant haemodynamic changes in heart failure patients. Against the background of continued high hospitalization and rehospitalization rates, 11 there is a clear need for improved diagnostic tools in the management of patients with chronic heart failure. The two studies in this volume add important information to the field of implantable monitoring by highlighting (i) the correlation between elevated cardiac filling pressures, pulmonary fluid retention, and decreasing intrathoracic impedance, (ii) the potential of further development in sensor technology for impedance monitoring, and (iii) the possibility of a multi-sensor approach to improve the diagnostic value of impedance monitoring in heart failure. Importantly, data from randomized clinical trials are warranted to demonstrate the effects of impedance monitoring on clinical endpoints such as death and heart failure hospitalization. However, this goal may not be achieved without better knowledge of how intrathoracic impedance data should be evaluated in the context of other haemodynamic and clinical information when patients are followed in clinical practice. Demonstrating correlations with cardiac filling pressures and extravascular lung water, the present papers provide additional support for the use of intrathoracic impedance monitoring in heart failure patients. This may help to overcome the resistance to routine applications of device-based monitoring that still exists. Conflict of interest: F.B. receives research funding by Medtronic, serves as a consultant to Biotronic and Boston Scientific and has received compensation for lectures by various pacemaker manufacturers.