| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical |
Department of Medicine, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong, China
Correspondence to: C.C. Szeto, Department of Medicine & Therapeutics, Prince of Wales Hospital, Chinese University of Hong Kong, Shatin, N.T., Hong Kong, China. ccszeto{at}cuhk.edu.hk
 |
ABSTRACT |
|---|
 TOP ABSTRACT PATIENTS AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
Objective: Cardiovascular disease (CVD) is the most
common cause of mortality in chronic peritoneal dialysis (PD) patients.
Increased arterial stiffness may be related to a high peritoneal permeability
resulting in fluid overload in PD patients. We studied the relations between
arterial stiffness, peritoneal transport, and radiographic parameters of
systemic fluid overload in a cohort of Chinese PD patients. Design: Prospective cohort study. Setting: University referral center. Patients: We studied 107 PD patients. Vascular pedicle
width and cardiothoracic ratio were measured from a plain postero-anterior
chest radiograph. Pulse wave velocity (PWV) was determined at
carotid–femoral (C-F) and carotid–radial sites. Peritoneal
transport was determined by the dialysate-to-plasma ratio (D/P) of creatinine
at 4 hours of dwell. Patients were followed for 9.4 ± 4.6
months. Outcome Measures: Duration of hospitalization;
actuarial and technique survival. Results: There were no relationships between
radiographic measures, arterial PWV, and D/P creatinine. However, both C-F PWV
and D/P creatinine were independent predictors of the number of
hospitalizations for CVD. None of the parameters correlated with mortality in
this study. Conclusions: There were no relationships between
radiological parameters of fluid overload, peritoneal transport
characteristics, and arterial PWV. Both C-F PWV and D/P creatinine were
independent predictors of the number of hospitalizations for CVD. Our result
suggests that arterial stiffness and high peritoneal transport each contribute
to the development of CVD in this group of patients.
KEY WORDS: Vascular pedicle width; cardiothoracic ratio; pulse wave velocity; peritoneal transport characteristics; cardiovascular disease.
End-stage renal disease is one of the most debilitating chronic medical illnesses. In Hong Kong there were more than 3500 patients on long-term dialysis in 2004 (1). Peritoneal dialysis (PD) is the preferred mode of renal replacement therapy in Hong Kong and accounts for 80% of patients requiring dialysis (1).
Cardiovascular disease (CVD) is the major cause of mortality and morbidity in PD patients (2,3). In addition to the classic risk factors of atherosclerosis such as diabetes, hypertension, and hyperlipidemia, uremia and possibly dialysis treatment per se play important roles in the pathogenesis of accelerated atherosclerosis in renal failure patients (4). It is now recognized that chronic intravascular hypervolemia may be a cause and consequence of arterial stiffness. Persistent excessive intravascular volume may cause remodeling of arterial structure (5,6). In turn, reduced arterial compliance increases left ventricular workload and reinforces fluid overload (7).
Peritoneal permeability plays a critical role in the success of PD: previous studies reported that patient survival was worse in PD patients with high peritoneal permeability (8,9). It is often postulated that high peritoneal transporters are prone to fluid overload, low serum albumin, and malnutrition (10), which may lead to excessive mortality. However, the factors that govern peritoneal transport are not fully understood. Several morphologic studies and mathematical models predict that capillary endothelium is the major site of resistance of peritoneal transport (11,12). Endothelial cells generate several substances that gradually change the function and structure of arterial wall, finally leading to arterial stiffness (13–19). Arterial permeability is one of the major determinants of the peritoneal small solute transport rate (13,14). Arterial stiffness is thought to increase peritoneal permeability by increasing arterial permeability.
In light of the available evidence, the aim of this study was to find whether there is any relationship between arterial stiffness, peritoneal transport characteristic, and intravascular volume status and whether these parameters might predict clinical outcome of PD patients.
|
PATIENTS AND METHODS |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
MEASUREMENT OF CHEST RADIOGRAPH
Postero-anterior chest radiographs were obtained in all patients. All
radiographic examinations were performed with computed radiography equipment
(Mobilett Plus; Siemens Medical Solutions, Malvern, PA, USA) using a
standardized technique (75 kV, 4 mAs, 180-cm film–focus distance; broad
tube focus). The images were assessed using a PACS (MagicView, model VA22E;
Siemens Medical Solutions) viewer (2 K monitor).
The method for measuring vascular pedicle width (VPW) has been described previously (20–23). Briefly, the right border of the VPW was the point at which the superior vena cava crossed the right main bronchus. The left border was the point of the subclavian artery exiting the aorta. The VPW was defined as the horizontal distance measured between the two points. Cardiothoracic ratio (CTR) was determined by the Danzer method (24).
PULSE WAVE VELOCITY (PWV) STUDY
Pulse wave velocity, an index of aortic stiffness, was measured using an
automatic computerized recorder and the results were analyzed using the
Complior SP program (Artech Medical, Pantin, France). The method of measuring
PWV has been described previously
(25). Briefly, within 1 week
of the PET, pressure-sensitive transducers were placed over the neck (carotid
artery), wrist (radial artery), and groin (femoral artery), with the patient
in the supine position. The PWV of the carotid–femoral (C-F PWV) and
carotid–radial (C-R PWV) territories were calculated by dividing the
distance between the sensors by the time corresponding to the period
separating the start of the rising phase of the carotid pulse wave and that of
the femoral and also the radial pulse waves. All PWV measurements were
performed by one observer; the intraobserver coefficient of variation was
0.118% – 0.609%.
PERITONEAL EQUILIBRATION TEST
We used the standard PET as described by Twardowski et al.
(26). All patients were in a
euvolemic state during the PET. Drainage and ultrafiltration volumes at 4
hours were documented. Dialysate-to-plasma ratios (D/P) of creatinine at 0, 2,
and 4 hours were calculated after correcting for glucose interference
(27). Mass transfer area
coefficients of creatinine normalized for body surface area were calculated by
the formula described by Krediet et al.
(28). The results of the PETs
were plotted on a PET graph and patients were classified as high,
high-average, low-average, or low transporters
(26).
DIALYSIS ADEQUACY AND NUTRITIONAL STATUS
Subjective Global Assessment and comprehensive
malnutrition–inflammation score were performed at enrolment. The 4-item
7-point scoring system, which has been validated in CAPD patients
(29), was used. Calculation of
the malnutrition–inflammation score was described previously
(30). Briefly, the
malnutrition–inflammation score consists of 4 main parts and 10
components, all scored from 0 (normal) to 3 (very severe). The total score
ranges from 0 to 30.
Dialysis adequacy by 24-hour dialysate and urine collections was determined. Total Kt/V was determined by standard methods (31). Residual glomerular filtration rate was calculated as the average of 24-hour urinary urea and creatinine clearances (32).
Fat-free edema-free body mass was measured by creatinine kinetics according to the formula of Forbes and Brunining (33). Normalized protein nitrogen appearance was determined by Bergström et al.'s formula (34) and normalized by ideal body weight.
OUTCOME MEASURES
Clinical outcomes included number of hospitalizations for CVD, total number
of hospitalizations for any cause, duration of hospitalizations, actuarial
patient survival, and technique survival. The causes of hospitalization were
divided into three areas: CVDs, infection, and other or unknown causes.
Cardiovascular diseases included cerebrovascular disease, coronary heart
disease, congestive heart failure, and peripheral vascular disease.
Transplantation, conversion to hemodialysis, and transfer to other units were
considered censored data for actuarial survival. The definition of technique
survival was patients were alive and on PD.
STATISTICAL ANALYSIS
Statistical analysis was performed using SPSS 13.0 software (SPSS Inc.,
Chicago, IL, USA). Results are presented as mean ± SD unless otherwise
stated. Comparisons between groups were performed using the unpaired Student's
t-test. Correlations are expressed by Pearson's or Spearman's correlation
coefficient, as appropriate. A p value < 0.5 was considered
statistically significant. All probabilities were two-tailed.
Survival rates were analyzed using Kaplan–Meier survival curves. The Cox proportional hazards model was used to identify independent predictors of actuarial survival and technique survival. Baseline variables, VPW, CTR, and PWV were added into the model. Backward stepwise elimination was applied to remove the insignificant variables.
Multivariate regression analysis was used to analyze the duration of hospitalization. Since the data were highly skewed, a log-linear regression model was used for analysis. In addition to VPW, CTR, C-R PWV, C-F PWV, and D/P creatinine at 4 hours, the models were constructed by age, diabetic status, Charlson comorbidity score, serum albumin, total Kt/V, normalized protein nitrogen appearance, fat-free edema-free body mass, and residual glomerular filtration rate. These parameters were selected for the construction of the models because of their importance in determining the clinical outcome of PD patients according to previous studies.
|
RESULTS |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
|
|
RELATIONS BETWEEN RADIOGRAPHIC MEASURES, PWV, PERITONEAL TRANSPORT, AND OTHER BASELINE PARAMETERS
There was no relationship between any chest radiograph measure and arterial
PWV (Table 3). Peritoneal
transport characteristic, as represented by D/P creatinine at 4 hours, also
did not correlate with VPW (r = –0.001, p = 0.9), CTR
(r = 0.146, p = 0.2), C-R PWV (r = 0.004,
p = 0.9), or C-F PWV (r = –0.013, p = 0.9).
However, VPW did correlate with body height (r = 0.386, p
< 0.001), body weight (r = 0.312, p = 0.003), and
high-density lipoprotein (HDL; r = –0.296, p = 0.005).
CTR correlated with age (r = 0.258, p = 0.014), body height
(r = –0.252, p = 0.016), systolic blood pressure (BP;
r = 0.227, p = 0.031), and pulse pressure (PP; r =
0.269, p = 0.01). C-F PWV correlated positively with age (r
= 0.354, p < 0.001), systolic BP (r = 0.429, p
< 0.001), PP (r = 0.488, p < 0.001), mean BP
(r = 0.318, p = 0.001), Charlson comorbidity score
(r = 0.587, p < 0.001), total cholesterol (r =
0.227, p = 0. 021), and triglyceride (r = 0.480, p
< 0.001), and correlated negatively with HDL (r = –0.225,
p = 0.022). C-R PWV correlated with body height (r = 0.228,
p = 0.018), diastolic BP (r = 0.341, p < 0.001),
and mean BP (r = 0.247, p = 0.01). D/P creatinine correlated
positively with malnutrition–inflammation score (r = 0.317,
p = 0.041) and negatively with serum albumin (r =
–0.455, p < 0.001), calcium (r = –0.237,
p = 0.021), and phosphate (r = –0.255, p =
0.013).
|
RELATION WITH HOSPITALIZATION
The patients were followed for 9.4 ± 4.6 months. There were 1001
days of hospitalization during the study period; 56 (52.3%) patients required
admission to hospital; 16 patients (15.0%) were admitted to hospital for CVD.
Multivariate regression analysis by log-linear modeling showed that the
independent predictors of the number of hospitalizations for CVD were C-F PWV
and D/P creatinine (Table 4).
In contrast, only the Charlson comorbidity score was an independent predictor
of total number of hospital admissions
(Table 4). Relative risk of an
individual hospitalization predictor is expressed as exponential coefficient
(ecoef).
|
RELATION WITH MORTALITY
During the study period there were 7 deaths, 2 patients had a kidney
transplant, 1 patient changed to hemodialysis due to severe peritonitis, and 2
patients transferred to other units. Cause of death was CVD (1 case),
peritonitis (1 case), non-peritonitis infection (1 case), and others or
unknown (4 cases). At 12 months the actuarial patient survival rate was 91.0%
and technique survival rate was 85.8%. By the Cox regression model, none of
the chest radiograph measures, arterial PWV, or peritoneal transport
characteristics were related to patient survival.
|
DISCUSSION |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
Contrary to our prediction, the present study did not show any relationships between arterial PWV (a marker of arterial stiffness), peritoneal transport characteristic, and radiograph measurements (surrogate markers of systemic fluid overload). In contrast, previous studies by Wang et al. (13,14) showed a significant relationship between C-F PWV and D/P creatinine. The value of C-F PWV in their study was 11.1 ± 2.3 m/s, which is substantially higher than in our cohort (9.93 ± 2.11 m/s). We are not sure of the exact reason for this discrepancy. It is, however, interesting to note that the studies by Wang et al. (13,14) recruited subjects from the northern part of China, which is an area of substantially higher dietary sodium intake compared to our local population. In our center, average sodium removal (dialysis plus urinary) in our PD patients is 1.15 ± 1.47 g/day (Szeto CC, unpublished data), compared to 2.3 – 3.2 g/day in PD patients from Beijing (36).
In the present study we did not find any relationship between peritoneal transport characteristics and VPW or CTR. Our result, however, does not imply there is no correlation between high transport and fluid overload. A previous study by Konings et al. (16) showed that the relationship between peritoneal transport and fluid overload was apparent only after a longitudinal study. In addition, a recent study by Cheng et al. (37) showed there is a negative relationship between volume overload and endothelial dysfunction, which is the major site of resistance to peritoneal transport (11,12). Furthermore, the reliability of VPW and CTR as indicators of intravascular volume in dialysis patients remains to be determined. Average VPW was 50.2 ± 5.8 mm in our study, which is within the normal range according to Ely et al. (20), indicating that most of our patients were not in gross fluid overload. The exact normal range of VPW in Asian populations is unknown.
We found that C-F PWV and D/P creatinine are independent predictors of hospitalization for CVD. A previous study (38) showed transport characteristic is not a predictor of hospitalization. To the best of our knowledge, this is the first study to show that C-F PWV and D/P creatinine are predictors of hospitalization for CVD at the same time. Since there is no internal correlation between these two parameters, our result suggests that arterial stiffness, which indicates the severity of atherosclerosis, and high peritoneal transport status, which predicts fluid overload, each contribute to the development of CVD in this population.
Some limitations of our study need to be noted. First, our patients were all in stable status without clinical features of gross fluid overload. As a result, the percentage of patients with abnormal VPW was small, which may have disguised any potential relationship. Second, the duration of follow-up time was short and our study certainly does not have the power (either sample size or duration of follow-up) to ascertain the relationship of PWV or peritoneal transport to patient survival.
In conclusion, there were no relationships between radiological parameters of fluid overload, peritoneal transport characteristic, and arterial pulse wave velocity. In the present study, both C-F PWV and D/P creatinine were independent predictors of the number of hospitalizations for CVD.
|
DISCLOSURE |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received 9 December 2008; accepted 25 March 2009.
|
REFERENCES |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
