Group 1 consisted of 151 consecutive outpatients (age 48.7 ± 12 years; 110 men) with end-stage DCM who fulfilled the inclusion criteria and were studied between 07/2001 and 12/2008 out of a mean of 182 patients/year referred for transplantation to the Deutsches Herzzentrum Berlin. Treatment included angiotensin-converting enzyme, diuretics, digoxin and beta-blockers in accordance with our institutional strategy. The results of the study did not influence the mode of treatment. Inclusion criteria were as follows: sinus rhythm, normal body mass index, no calcification or stenosis in the carotid artery studied, and mitral regurgitation of less than grade 1. The follow-up period was 31 ± 8 months.

Group 2 consisted of 11 consecutive patients (age 50 ± 11 years; 6 men) with sinus rhythm and “true decompensation”, admitted for assist device implantation. This group is characterized by frank pulmonary decompensation and is qualified as Group “D” according to the ACC/AHA guidelines. The patients were studied before large doses of catecholamines were used to stabilize their hemodynamic condition. However, during the course of treatment decompensation appeared to be refractory to catecholamines and no medical means were able to keep the patients alive. All had to be treated by implantation of a ventricular assist device (VAD) to save their lives. The baseline characteristics of the patients are shown in Table 1. There were no demographic or somatic (weight and height) differences between the groups. Sinus rhythm was an inclusion criterion for the study. A maximal oxygen consumption study was performed during the same investigations only in Group 1.

The invasive wave intensity (WI) index is well described by Parker et al. [11] and was primarily calculated from invasive parameters as the product of the derivatives of velocity inside the artery and the pressure of the same artery. Non-invasive wave intensity (NWI) is defined on the basis of echocardiography (ALOKA ultrasonic system, Tokyo, Japan) and uses the protocol proposed by Sugawara et al. [12] and Harada et al. [10,12]. The pressure waveform is derived from the diameter change curve of the arterial wall. Assessment of the diameter curve is based on echo tracking beam configuration technology with independent beam steering of the carotid artery. Waveform (dU) is calculated in real time as the difference between posterior and anterior wall displacement during short time intervals (5 ms) (Figure 1) (ALOKA, Tokyo, Japan). The waveform (dU) closely corresponds to pressure change (dP) in a carotid artery measured invasively with high correlation quotient (goodness of fit r² = 0.97) and can be substituted for the invasively recorded pressure waveform [12]. Velocity measurements were based on 5 MHz continuous wave Doppler integrated equipment inside the studied artery at the same time as diameter changes. The formula for NWI is as follows:

NWI = (dU/dt) • (dV Doppler/dt)

dU—difference in displacement of carotid artery; V—flow velocity measured by Doppler.

The NWI index materializes as the curve consisting of two mathematically defined peaks (1st and 2nd). The 1st peak is the “compression wave” and is mathematically positive as a product of two positive values of pressure (green curve, Figure 2) and flow waveform (red curve). The compression wave describes the volume of the blood ejected in systole from the left ventricle conduit (aorta and arterial vessels). The 2nd peak of WI (“expansion wave”) is calculated in the same way as the 1st peak, which means that it is the result of algebraically calculated pressure and flow inside the vessel after ejection. While both curves (pressure and flow) in this phase are negative, algebraically their derivates produce positive peak of WI. The WI curve is recorded in yellow. The height of the 1st peak depends on the value of pressure and flow and their dynamics at the beginning of ejection (Figure 2).

All patients received serial echocardiography, and were followed up according to our institutional protocol published elsewhere [13]. In patients who were hemodynamically instable echocardiography was performed daily (Group 2).

Pulsed Doppler mitral flow analysis. In each patient, LV diastole flow velocity waves from five cardiac cycles were recorded and averaged. The following measurements were obtained: peak velocity of early diastolic filling wave (E); peak velocity of late filling (A).

Values of 1st and 2nd peak calculated in the truly decompensated group with added standard deviation (SD) were used as reference values to identify the risk of true decompensation (cut-off for 1st peak value at 4,100 mmHg*s³). For the purposes of this analysis these values were accepted for identification of patients at risk for “events” (death, true decompensation requiring assist device support or urgent transplantation).

Demographic data and other characteristics are shown as number, as mean or as median ± SD.

The hypothesis was that the hemodynamic deterioration leads to three clinical situations, death, assist device implantation or urgent heart transplantation, and these situations were primary endpoints (events for statistical purposes).

The second hypothesis was set that the value of 1st and 2nd WI peak assessed in the group of critically ill patients (n = 11) during true decompensation could be used as a “value of risk for true decompensation” in the ambulatory patients and would statistically correspond with the clinical endpoints of ambulatory patients.

A cumulative survival curve was created to express the results of follow-up. Statistical significance was assumed for P values

The baseline clinical characteristics of the two groups of patients studied are outlined in Table 1. The patients do not vary in terms of the duration of congestive heart failure symptoms or number of decompensations. The height and weight did not differ statistically between the groups.

The mean left ventricular filling pressure and mean pulmonary artery pressure (PAP) in Group 2 (during decompensation) were significantly lower compared to those of Group 1 and were 18 ± 9.81 vs. 27 ± 11 (pvs. 28.6 ± 6 mmHg (pvs. 2.0 ± 0.3 (p

The stable ambulatory group of patients (Group 1) compared to Group 2 did not have significantly smaller median LV diameter: 72 (min. 61, max. 100) mm vs. 75 (min. 60, max. 90) mm and median EF was also similar: 20 (min. 10, max. 40)% vs. 19 (min. 15, max. 25)% (Table 2).

The differences between groups mentioned above were to be expected by the design of the study. The data were obtained from patients during frank decompensation (Group 2) where the hemodynamics per se was characterized by high filling pressures. The LV dimensions and systolic function (EF) did not differ significantly between Groups 1 and 2 and were: 72.0 (51–100) vs. 75.0 (60–90) mm, p = 0.43, and 20.0 (10–40) vs. 1 .0 (15–25)%, p = 0.1, respectively.

Non-invasive WI data obtained from the carotid artery including the values of 1st and 2nd WI peak and ß value are listed in Table 2. The truly decompensated group (Group 2) had significantly lower 1st peak than the stable patients (Group 1): 2.900 ± 900 vs. 5400 ± 3.500 mmHg*s³ (p

Patients in Group 1 (n = 151) during a mean follow-up period of 31 ± 8 months experienced the following events which were defined as end-points for statistical purposes: 15 cardiac deaths (9.9%) before a VAD could be implanted to save the patients’ lives and 10 (6.6%) VAD implantations (true decompensation where stabilization by clinical means was not possible); 19 (12.6%) other patients received heart transplantation because of clinical deterioration. On the basis of these findings the patients were divided into two groups: event-free, Group A (n = 107) (stable ambulatory patients during follow-up) and event-positive, Group B (n = 44). These groups of patients were extensively studied to identify prognostic risk factors (Table 3).

Invasively assessed systolic pulmonary artery pressure (PAP) did not vary significantly between Groups A (event-free) and B (event-positive); however the diastolic PAP and left filling pressure were elevated in the event-positive group: median 18 (14–45) vs. 21 (13–53) mmHg and 17.0 (4–36) (p = 0.069) vs. 19.0 (7–40) mmHg (p = 0.058), respectively. The difference in stroke volume (SV) between the two groups did not reach statistical significance; the median was 47.1 (19–108) mL vs. 39.5 (7–40) mL (p = 0.042), respectively.

There were no significant differences between the event-free and event-positive groups regarding echocardiographic parameters (EDD, EF and mitral flow pattern). Median EF was identical in the two groups (20%).

The 44 patients classified as having “events” (29.5%) and forming Group B were removed from follow-up to assess freedom from event and construct the survival curve (Figure 3).

The mean values assessed in the carotid artery for NWI calculation were as follows:

heart rate and maximal pressures (systolic and diastolic) as well as diastolic minimal flow did not differ significantly between Group A and B; however the median value of maximal velocity in the carotid artery was significantly higher (p = 0.057) in the event-free patients (Group A) than in the “event” group (B) and was respectively 0.626 (0.230–1.469) and 0.524 m/s (0.216–1.1215) (Table 3).

Group A was characterized by 1st peak WI which was higher (6.400 ± 3.300) than the “value of risk for true decompensation” (4,100 mmHg*s³). Patients in whom the assessed 1st peak was small (

The group of 11 consecutive patients (age 50 ± 11 years; 6 men) possessing sinus rhythm and being “truly decompensated” during chronic disease (duration of congestive symptoms was 5.9 ± 0.9 years with a mean of 1.6 ± 0.9 episodes of frank decompensation) were admitted for assist device implantation. We investigated the patients in hemodynamic cardiovascular shock but before norepinephrine was used; it had not been possible to stabilize them by medical means. This group was chosen as candidates for establishing “critical values” of peak 1st and 2nd and ß value of WI.

Although NWI was assessed in the carotid artery it defines cooperation of the heart with the arterial conduit (aorta and arterial system) as a whole and not exclusively the arterial system of the head. This is the law of “connected compartments” (vessels). It seems to be true that for patients in whom heart dysfunction is present extracardiac factors may play an important role, leading to progressive decompensation; elevated filling pressure may cause heart dysfunction but may not always be the result of the loss of myocardial performance [17]. Newly published data supports the hypothesis that “hypervolemic episodes” in patients suffering from chronic heart failure could be the cause of decompensation events [18] rather than primary myocardial failure. This supports the results of our study that cooperation of the failing heart and arterial conduit can be an important factor for outcome prediction in patients suffering from congestive heart failure. Indeed, our prospective study demonstrates that 1st peak of WI which defines the systolic interaction of the arterial and ventricle system was highly significant for survival prediction if the 2nd peak was preserved. Preserved 2nd and low 1st peak has to be explained as a result of poor systolic function of the left chamber (small stroke volume and poor ejection dynamics) and poor response of the arterial system to ejected blood. Some research indicates that diastolic function of the heart may be reflected in the 2nd peak [19]; this could also be interpreted as high resistance of the vessel conduit causing diastolic dysfunction. It is believed that interaction or coupling is dependent on the “vigor” of the volume ejection into the aorta, producing the pressure wave equilibrium phenomenon forwards and backwards as well as the flow forwards and backwards inside the artery. Arterial stiffness factor ß was found to be also a good predictor of mortality in our group of patients but was not as prominent as 1st peak. High ß values (above 12) indicated patients at risk of death. There is no clear explanation of this phenomenon. However there were no differences in ß value between stable ambulatory patients and the truly decompensated patients (Table 1).

It should be kept in mind that the arterial tree functions as a “conductor” and plays a vital role as an active transmitter of blood supply to the peripheral tissue. Whether failure of this interaction has been provoked primarily by the steady failing of severely diseased myocardium or primarily by failure of the arterial tree could not be clarified; this was not the aim of our study. This study not only shows NWI to be a new clinical marker that is important for follow-up but also opens up new possibilities of using this new instrument to influence the medical treatment and the selection of specific treatment options. The reactivity of the arterial system to any kind of physiological stimulation is well known from animal studies [20]. However, patients suffering from congestive heart failure in general and specially during the advanced stages of the disease are very stable and do not demonstrate variability of WI parameters from beat to beat. This method was validated and demonstrated good reproducibility in human beings [21].