Author + information
- Received August 10, 1998
- Revision received November 6, 1998
- Accepted December 23, 1998
- Published online April 1, 1999.
- Cheol Whan Lee, MD∗,
- Jung-Hee Lee, PhD†,
- Jae-Joong Kim, MD∗,
- Seong-Wook Park, MD, PhD, FACC∗,
- Myeong-Ki Hong, MD∗,
- Sang-Tae Kim, BS†,
- Tae-Hwan Lim, MD, PhD‡ and
- Seung-Jung Park, MD, PhD, FACC∗,* ()
- ↵*Reprint requests and correspondence: Dr. Seung-Jung Park, Department of Medicine, University of Ulsan, Asan Medical Center, 388-1 Poongnap-dong, Songpa-gu, Seoul, 138-736, South Korea
Using proton magnetic resonance spectroscopy, we investigated cerebral metabolism and its determinants in congestive heart failure (CHF), and the effects of cardiac transplantation on these measurements.
Few data are available about cerebral metabolism in CHF.
Fifty patients with CHF (ejection fraction ≤35%) and 20 healthy volunteers were included for this study. Of the patients, 10 patients underwent heart transplantation. All subjects performed symptom-limited bicycle exercise test. Proton magnetic resonance spectroscopy (1H MRS) was obtained from localized regions (8 to 10 ml) of occipital gray matter (OGM) and parietal white matter (PWM). Absolute levels of the metabolites (N-acetylaspartate, creatine, choline, myo-inositol) were calculated.
In PWM only creatine level was significantly lower in CHF than in control subjects, but in OGM all four metabolite levels were decreased in CHF. The creatine level was independently correlated with half-recovery time and duration of heart failure symptoms in PWM (r = −0.56, p < 0.05), and with peak oxygen consumption and serum sodium concentration in OGM (r = 0.58, p < 0.05). Cerebral metabolic abnormalities were improved after successful cardiac transplantation.
This study shows that cerebral metabolism is abnormally deranged in advanced CHF and it may serve as a potential marker of the disease severity.
Congestive heart failure (CHF) is a common, disabling disorder with high morbidity and mortality. A complex series of compensatory mechanisms occurs to maintain perfusion to vital organs in response to the progressive reduction of cardiac output (1–5). The brain is a critical organ, but it should also adjust to this change as heart failure becomes chronic. Cognitive dysfunction is a common problem experienced by patients with CHF (6–10). It may result from adaptation failure of the brain, which leads to poor quality of life and eventually to death. Therefore, cerebral adaptation mechanisms certainly deserve to be considered in clinical assessment of CHF patients.
Proton magnetic resonance spectroscopy (1H MRS) is a sensitive technique to detect metabolic changes of the brain (11). For this reason, 1H MRS has been convincingly used to assess metabolic changes of the brain in various clinical diseases (11–17). Cerebral metabolism is presumably abnormal due to overall and local disturbances of cerebral blood flow in CHF (18,19). However, there is a lack of information regarding the cerebral metabolism in CHF. We assumed that cerebral metabolism is abnormally deranged, and it may represent the disease severity in CHF.
To address this hypothesis, we investigated cerebral metabolism and its determinants in patients with advanced CHF and examined the effects of cardiac transplantation on these measurements using 1H MRS examination.
Fifty consecutive patients with advanced CHF (left ventricular ejection fraction ≤35%) and 20 healthy volunteers were prospectively included for this study. All patients were clinically stable with medical treatment in the last four weeks; 50 patients were treated with digoxin, 48 patients with angiotensin-converting enzyme inhibitors and 46 patients with diuretics. Of the patients, 10 patients underwent cardiac transplantation. Patients were excluded if they had cerebrovascular disease, history of drug or alcohol abuse, chronic liver disease, chronic renal failure or any other diseases affecting cognitive function. The study was approved by the ethics committee of the Asan Medical Center, and no adverse events were recorded in any individuals undergoing the protocol. All subjects gave written informed consent to participate in the study.
Clinical and laboratory evaluation
All subjects were evaluated using a standardized protocol including patient history, physical examination and laboratory studies on the day of 1H MRS examination. New York Heart Association (NYHA) functional class was also determined on the day of 1H MRS examination.
Echocardiography and multigated blood pool scan
Two-dimensional Doppler echocardiography was performed with standard techniques in all subjects, and left ventricular ejection fraction was measured using multigated blood pool scan.
Exercise test was performed in all subjects on an upright electromagnetically braked bicycle (MedGraphics, Cardiopulmonary Diagnostic Systems, St. Paul, Minnesota) using a ramp protocol (10-W ramp with an increment of 10 W/min in CHF patients; 20-W ramp with an increment of 20 W/min in normal control subjects) in the morning after a light standard meal. Exercise test was terminated at a point when subjects were unable to continue due to dyspnea or fatigue. Subjects remained on the bicycle after the exercise during the recovery phase until the respiratory exchange ratio was less than 1 for 20 s. Respiratory gas analysis was carried out with a computerized system (MedGraphics, Cardiopulmonary Diagnostic Systems, St. Paul, Minnesota), calibrated with standard gas of known concentration before each test. The peak oxygen consumption was defined as highest 15-s average oxygen consumption. Recovery kinetics are characterized by measuring the half-recovery time, which is the time required for a 50% fall in the peak value (20,21).
Proton magnetic resonance spectroscopic examination
All subjects underwent 1H MRS examination of the brain, and patients receiving cardiac transplantation were asked to have a follow-up examination after two months. Localized in vivo 1H MRS was performed on a GE 1.5-T SIGNA system equipped with shielded gradients (General Electric Medical Systems, Milwaukee, Wisconsin). Proton magnetic resonance spectroscopy was performed following a T1-weighted magnetic resonance imaging. Image-guided stimulated echo acquisition mode spectra were obtained in two locations (voxel volume, 8 to 10 ml), a mostly parietal white matter (PWM) voxel and a mostly occipital gray matter (OGM) voxel. The spectral acquisition parameters were echo time of 30 ms, repetition time of 3 s and 36 averages with PROBE (proton brain examination) (General Electric Medical Systems, Milwaukee, Wisconsin). A three-pulse chemical shift selective sequence was used for suppression of the water signal. All raw PROBE data were transferred to a Sun Spark-10 workstation (SUN Computer, Sunnyvale, California) and processed according to the method described by Kreis et al. (12). Peaks were identified with known chemical shifts: N-acetylaspartate (NAA), 2.02 ppm; creatine (Cr), 3.03 ppm; choline (Cho), 3.22 ppm (choline-containing compounds), and myo-inositol (mI), 3.56 ppm. The absolute concentrations of the cerebral metabolites were calculated using the brain water signal as an internal reference from the PROBE data (22–24)and expressed in mmol/kg of wet weight. The reference brain water signals used in our calculation were 64% in PWM and 76% in OGM, as calculated according to the method described by Ernst et al. (23)for normal volunteers. All spectra were reviewed blindly to the clinical data.
Statistical analysis was performed using SPSS 7.5 for Windows. Data were expressed as mean ± SD for continuous variables, and frequencies for the categorical variables. Continuous variables were compared by unpaired Student ttest, categorical variables by chi-square test and serial changes of cerebral metabolites before and after cardiac transplantation by paired Student ttest. Linear regression analysis was performed on all variables to identify determinants of the cerebral metabolite levels, and variables found to be significant by univariate analysis were entered into a stepwise multiple linear regression analysis to determine their independent relationship to the cerebral metabolite levels. A p value < 0.05 was considered statistically significant.
The baseline clinical and laboratory findings are summarized in Table 1. The causes of CHF were dilated cardiomyopathy in 45 patients, valvular heart disease in three patients and ischemic heart disease in two patients. Age, gender and mean blood pressure were similar between the two groups. However, there were significant differences in heart rate, NYHA functional class, rhythm status, left ventricular function or dimension, exercise capacity and serum sodium or osmolar concentrations between the two groups.
Magnetic resonance image
The brain was normal in all study subjects on magnetic resonance imaging. Cerebral atrophy, cerebral infarct or tumors were not seen in any subjects included for this study.
Proton magnetic resonance spectroscopy
Metabolic changes in PWM
Typical 1H MRS spectra obtained from a healthy volunteer and a patient with CHF are shown in Figure 1, which represent a markedly decreased creatine peak, but normal peaks of other metabolites in the CHF patient compared with those in the normal control subject. The creatine level in PWM was significantly decreased in CHF compared with that in the normal control subjects, but levels of other metabolites (NAA, Cho and mI) were not different between the two groups (Table 2). Factors associated with a decreased level of creatine were half-recovery time, peak oxygen consumption, duration of heart failure symptoms, left ventricular function and NYHA functional class (Table 3). However, half-recovery time and duration of heart failure symptoms were independently associated with creatine level by multivariate regression analysis (r = −0.56, p < 0.05). Interestingly, creatine level was significantly lower in patients with heart failure duration more than six months (n = 18) than in patients with duration less than six months (6.44 ± 0.90 mmol/kg vs. 6.94 ± 0.60, respectively, p < 0.05).
Metabolic changes in occipital gray matter
In contrast to PWM, creatine, choline and myo-inositol levels in OGM were markedly decreased in CHF compared with those in normal control subjects (Fig. 2, Table 2). In addition, NAA level in OGM was also decreased in CHF (Table 2). The creatine level was significantly related to peak oxygen consumption, half-recovery time, serum sodium or osmolar concentration and NYHA functional class (Table 3). However, peak oxygen consumption and serum sodium concentration were independent predictors of the occipital Cr level by multivariate regression analysis (r = 0.58, p < 0.05). N-acetylaspartate level was associated with serum sodium concentration (r = 0.34, p < 0.05) and peak oxygen consumption (r = 0.30, p < 0.05). In the multivariate analysis, however, only serum sodium concentration remained a significant predictor of NAA level. The Cho level was independently correlated with serum sodium concentration and NYHA functional class (r = 0.47, p < 0.05) and the myo-inositol level with serum osmolarity (r = 0.42, p < 0.01).
Effect of cardiac transplantation on cerebral metabolites
Figure 2shows characteristic change of creatine peak in PWM (a) and OGM (b) after successful cardiac transplantation, and the serial changes of all metabolite levels before and two months after successful cardiac transplantation are listed in Table 4. The levels of Cr/Cho in PWM, and of NAA and Cr in OGM were significantly improved after successful cardiac transplantation (p < 0.05). Myo-inositol and choline levels in OGM were also increased after successful cardiac transplantation, but this did not achieve statistical significance.
In the present study we demonstrate that cerebral metabolism is abnormally deranged in patients with advanced CHF, but improves after successful cardiac transplantation. Furthermore, we show that the spectroscopic abnormalities are related to exercise capacity and duration of heart failure symptoms, thereby serving as a potential marker for the disease severity. These findings may provide a neurochemical basis for the understanding of cognitive dysfunction in advanced CHF (6–10).
Proton magnetic resonance spectroscopy is a highly sensitive and reliable method that allows us noninvasively to investigate metabolic changes in a well localized region of the brain in vivo and, therefore, provides quantitative information on regional cerebral metabolism. In most clinical studies the results of 1H MRS have been reported as metabolite ratio measurements, as opposed to quantitation of metabolite levels (11–17,25). Creatine has been used as an internal standard assuming that the creatine level is constant in the brain. However, its levels can be changed according to the nature of clinical diseases as shown in this study. Furthermore, metabolite ratio measurements may be misleading in cases in which all metabolites change simultaneously in the same direction. In contrast, quantitation of absolute metabolite levels overcomes this limitation. For these reasons, we analyzed all data in terms of absolute metabolite levels.
Metabolic changes of parietal white matter
Creatine level in PWM was remarkably decreased in CHF, but recovered after successful cardiac transplantation. In contrast, levels of other metabolites (N-acetylaspartate, choline and myo-inositol) were not different between the two groups. These findings suggest that patients with CHF have a cerebral energy deficit in the PMW region because creatine is critical for cerebral energy metabolism and reserve. Cerebral metabolite levels may be affected by osmolar status (26). However, myo-inositol level, a marker of reversible cerebral osmolyte, was not significantly changed in CHF. Furthermore, serum sodium concentration was mildly decreased in most of the patients. Therefore, we suppose that decreased creatine level in PWM may be a characteristic 1H MRS pattern of CHF, representing cerebral energy deficit. The lactate peak, an end product of anaerobic glycolysis, was not observed in this study, which may represent metabolic compensation of the brain to chronic cerebral hypoperfusion. Interesting, and somewhat unexpected, was an increase in parietal Cho level two months after successful cardiac transplantation. The mechanism underlying this puzzling overshoot phenomenon remains unclear, which requires further studies.
Creatine level in PWM was correlated with exercise capacity and duration of heart failure symptoms. Exercise capacity has been demonstrated to be impaired in parallel with the severity of the disease, and was reported to be an important determinant of prognosis in patients with CHF (1–3,20,21). In these regards, decreased level of creatine in PWM region seems to reflect chronicity of the disease, and severity of left ventricular dysfunction in patients with CHF.
Metabolic changes of occipital gray matter
Metabolites in gray matter were more dramatically changed than those in PWM. These may result from regional differences of susceptibility to chronic cerebral hypoperfusion and/or osmolar changes. Metabolic changes were more closely related to serum osmolarity in gray matter than in white matter. However, not all of the decreased level in metabolites can be explained by the changes in serum osmolarity, because hyposmolar state was mild in most of the patients, and the creatine levels were also independently correlated to peak oxygen consumption. In general, metabolic abnormalities appear to develop earlier, and progress more rapidly, in gray matter than in white matter. However, it remains to be elucidated whether diffuse metabolic abnormality of the brain is a distinct occipital 1H MRS pattern in CHF.
N-acetylaspartate was suggested as a neuronal marker, loss of NAA being generally an accompaniment of neuronal loss (11). Until now, it is still controversial whether loss of NAA can be reversible according to the disease activity in a clinical setting. Interestingly, our study shows that it can be recovered after successful cardiac transplantation. However, this must be cautiously interpreted as signs of neuronal recovery because several other metabolites can contribute to the spectral region of NAA, such as glutamate, aspartate and others.
It is interesting that the 10 patients who received transplantation had lower cerebral metabolite levels compared with those of the total CHF patients before their transplantation. After transplantation, their metabolic level, in general, was closer to that of the total CHF patients than to that of the control group. Long-term follow-up data will be needed to conclude whether their metabolite levels are completely normalized or not.
Mechanisms of abnormal cerebral metabolism
At present, we are not able to explain exactly the pathophyisiologic mechanisms of cerebral metabolic abnormalities in advanced CHF. As a possible mechanism by which cerebral metabolism is deranged, chronic cerebral hypoperfusion may play a primary role (18,19). It is well known that regional blood flow is decreased and redistributed to major organs according to the demands in advanced CHF. This is achieved by a complex interaction between neural, humoral and local factors with a reduction of cardiac output. In general, depending on the severity of heart failure, cerebral and coronary blood flow is maintained at the expense of renal blood flow and skeletal muscle perfusion, because autoregulation is well developed in these organs. However, redistribution of regional blood flow cannot be reliably evaluated because it is difficult to assess blood flow to several organs simultaneously in human. We suppose that cerebral blood flow may decrease even under resting conditions in advanced CHF, and chronic cerebral hypoperfusion may induce metabolic abnormalities of the brain. In addition, it is likely that regional variations in the cerebral metabolites may reflect regional redistribution of cerebral blood flow and/or regional susceptibility to chronic cerebral hypoperfusion.
Until now, psychometric analysis has been used for evaluation of cognitive function in CHF, but it may not be objective and accurate, because it is affected by various biases (27,28). In contrast, 1H MRS may provide reliable and objective metabolic information for evaluation of cerebral function.
Cardiac transplantation is the only definitive treatment for patients with advanced CHF. However, assessment of prognosis in patients with CHF is very difficult because of the extremely variable clinical course. Until now, the exercise cardiopulmonary function test has been used to determine the timing of cardiac transplantation in ambulatory patients with advanced CHF (29,30), but it has many limitations, because exercise capacity can be influenced by numerous noncardiac factors and exercise training. We believe that 1H MRS may also be used to evaluate the disease severity and prognosis of patients in CHF. Previously, the role of skeletal muscle abnormalities and evidence of their role in exercise capacity have been extensively studied and reviewed in CHF (5). However, little interest has been directed toward adaptation mechanisms of the brain in CHF despite its clinical importance. Therefore, we propose that cerebral metabolism in CHF may be an attractive area for further investigation in the future, and it may provide important prognostic information. Further studies in larger patient populations may be needed to determine the role of 1H MRS as a screening test for patients with a worse prognosis.
Several potential limitations need to be addressed. First, psychometric analysis and biochemical tests were not performed in this study. Therefore, we could not evaluate the association between the cerebral metabolic abnormalities, biochemical variables of left ventricular dysfunction and cognitive scores by psychometric measures. Second, cerebral blood flow was not examined to investigate regional cerebral metabolism. Positron emission tomography may allow the study of blood flow in specific cerebral regions and, therefore, may be used in conjunction with 1H MRS to obtain simultaneous information on cerebral blood flow and cerebral metabolism. Third, most patients were taking standard medication for treatment of CHF and, therefore, drug effects on cerebral metabolites could not be evaluated. Finally, the regions of OGM may not exclude contamination from white matter or cerebrospinal fluid, because it is difficult to select 8 to 10 ml of pure OGM. However, placing localization voxel was carefully performed in the same region for all subjects to increase consistency of the region selected.
On the basis of our findings, we conclude that cerebral metabolism is abnormally deranged on 1H MRS in advanced CHF, and it may be used as a potential marker of the disease severity.
☆ This study was supported in part by The Korea Science and Engineering Foundation (KOSEF) during 1996–1998 (grant #961-0706-049-2) and by the Asan Institute for Life Science (grant #98-217).
- congestive heart failure
- 1H MRS
- proton magnetic resonance spectroscopy
- New York Heart Association
- occipital gray matter
- parietal white matter
- Received August 10, 1998.
- Revision received November 6, 1998.
- Accepted December 23, 1998.
- American College of Cardiology
- Costanzo M.R,
- Augustine S,
- Bourge R,
- et al.
- Clark A.L,
- Poole-Wilson P.A,
- Coats A.J
- Grimm M,
- Yeganehfar W,
- Laufer G,
- et al.
- Cohen-Solal A,
- Laperche T,
- Morvan D,
- et al.
- de Groote P,
- Millaire A,
- Decoulx E,
- et al.
- Launer L.J,
- Wind A.W,
- Deeg D.J
- Stelken A.M,
- Younis L.T,
- Jennison S,
- et al.
- Mancini D.M,
- Eisen H,
- Kussmaul W,
- et al.