 |
Periodicals Home | Search | User Pref |
Help | Logout |
| AJKD Home | Table
of Contents |
All Issues | Order | About this Journal |
<< Issue |
>> Issue |
|
Periodicals Home | Search | User Pref |
Help | Logout |
| AJKD Home | Table
of Contents |
All Issues | Order | About this Journal |
<< Issue |
>> Issue |
 |
Kamyar Kalantar-Zadeh, MD, MPHaSections |
|
| Abstract | TOP |
Background: Elements of malnutrition-inflammation complex syndrome (MICS) may blunt the responsiveness of anemia of end-stage renal disease (ESRD) to recombinant human erythropoietin (EPO).
Methods: The authors examined cross-sectional associations between the required dose of EPO within a 13-week interval as prescribed by practicing nephrologists who were blind to the study and several laboratory values known to be related to nutrition and/or inflammation, as well as the malnutrition-inflammation score (MIS), which is a fully quantitative assessment tool based on the subjective global assessment of nutrition.
Results: A total of 339 maintenance hemodialysis (MHD) outpatients,
including 181 men, who were aged 54.7 ± 14.5 years (mean ± SD), who had
undergone dialysis for 36.3 ± 33.2 months, were selected randomly from 7 DaVita
dialysis units in Los Angeles South/East Bay area. The average weekly dose of
administered recombinant human EPO within a 13-week interval was 217 ± 187 U/kg.
Patients were receiving intravenous iron supplementation (iron gluconate or
dextran) averaging 39.5 ± 47.5 mg/wk. The MIS and serum concentrations of
high-sensitivity C-reactive protein, interleukin 6 (IL-6), tumor necrosis
factor-
, and lactate dehydrogenase
had positive correlation with required EPO dose and EPO responsiveness index
(EPO divided by hemoglobin), whereas serum total iron binding capacity (TIBC),
prealbumin and total cholesterol, as well as blood lymphocyte count had
statistically significant but negative correlations with indices of refractory
anemia. Most correlations remained significant even after multivariate
adjustment for case-mix and anemia factors and other relevant covariates.
Similar associations were noticed across EPO per body weight tertiles via
analysis of variance and after estimating odds ratio for higher versus lower
tertile via logistic regression after same case-mix adjustment.
Conclusion: The existence of elements of MICS as indicated by a high MIS and increased levels of proinflammatory cytokines such as IL-6 as well as decreased nutritional values such as low serum concentrations of total cholesterol, prealbumin, and TIBC correlates with EPO hyporesponsiveness in MHD patients.
| (Click on a term to search this journal for other articles containing that term.) |
| Keywords: Recombinant
human erythropoietin (EPO), inflammation,
protein-energy
malnutrition, cytokines,
malnutrition-inflammation
complex syndrome (MICS), malnutrition-inflammation
score (MIS) |
| Methods | TOP |
Patients
The participating
subjects in the NIED Study are from a pool of approximately 1,200 MHD
outpatients in 8 DaVita, Inc, dialysis facilities in the South Bay Los Angeles
area. Dialysis units, each treating 105 to 187 MHD patients at the time of the
start of the study, are divided into 2 operational groups according to yearly
quarters ([1] October-April group and [2] January-July group). Each operational
group includes 4 dialysis units, of which approximately 180 consenting MHD
patients are evaluated semiannually, including 30 to 60 MHD patients who are
selected randomly at each dialysis facility (see NIED Study website at www.nephrology.rei.edy/NIED.htm
for more details). Inclusion criteria were those outpatients who were undergoing
MHD for at least 8 weeks, were 18 years or older, and who signed the written
consent form. Patients with a life expectancy of less than 6 months, for
example, because of a metastaic malignancy or terminal HIV disease, were
excluded. In the initial phase of the NIED Study (October 2001 through March
2002), 385 patients signed the written consent form. However, 1 dialysis
facility (in Norwalk, CA) with 30 participating subjects did not have complete
records of the administered EPO and iron doses during the designated study
period and hence was excluded from this analysis. Of the remaining 355 MHD
patients from 7 dialysis facilities, 339 consenting patients underwent the blood
tests and nutritional evaluations examined in this analysis, whereas 16 other
patients were not present in the dialysis units at the time of blood draws. The
medical chart of each MHD patient was reviewed by a nephrologist, and data
pertaining to underlying kidney disease, cardiovascular history, and other
comorbid conditions were extracted. A modified version of the Charlson
comorbidity index, ie, without the age and kidney disease components, was used
to assess the severity of comorbidity.18,19
Erythropoietin, EPO indices, and
iron
In all 7 dialysis facilities, precise documentation of
the administered doses of recombinant human erythropoietin or epoeitin (Epogen)
(EPO) and iron was available. The total dose of EPO (U/wk) among all 339 MHD
patients of this analysis was calculated over a 13-week interval, ie, from
October 1 through December 31, 2001, for group 1 and from January 1 through
March 31, 2002, for group 2. The average weekly EPO dose then was calculated by
dividing the total 3-month dose by 13. For those patients who missed more than 1
week of dialysis treatment or who left the cohort before the end of the third
month (because of death, transplant), the average EPO dose per week was
calculated using the actual numbers of weeks they contributed to the cohort. EPO
responsiveness (resistance) index was defined as the average weekly EPO dose
divided by average blood hemoglobin as described by Gunnell et al6
to normalize the amount of required EPO for the degree of severity of anemia.
Both average EPO dose and EPO responsiveness index also were divided by body
weight (in kilograms) to obtain 2 additional variables to indicate the required
EPO dose per kilogram of body weight. The route of EPO administration was
considered intravenous (IV) or subcutaneous (SQ) based on the administration of
more than 50% of EPO dose given within the study interval. All but 64 MHD
patients (19%) received EPO intravenously.
Of all 339 MHD patients who
were studied in this analysis, 217 patients received IV iron at least once
within a 3-month period including 103 patients who received iron gluconate
(Ferrlecit), 62.5 to 125 mg, and 14 patients (mostly those under the care of
Kaiser Permanente of Southern California) who received iron dextran (Infed), 50
to 100 mg weekly to monthly. No patient received iron sucrose (Venofer) during
this study period. Only 18 patients were on oral iron medications.
The
decision as to how much EPO and/or iron was to be administered to each study
patient was made blindly by at least 42 nephrologists who were in charge of
dialysis treatment care of these 339 MHD patients. Most nephrologists were not
aware of the periods in which this analysis was conducted. Almost all
nephrologists treated the anemia of their MHD patients according to Kidney
Disease and Dialysis Outcomes Quality Initiative (K/DOQI) guidelines,6
ie, to achieve a targeted hemoglobin of 11 to 12 g/dL (110 to 120 g/L) and/or a
hematocrit of 33% to 36%. Furthermore, 8 nephrologists were asked randomly if
they were aware of K/DOQI guidelines and whether they used these recommendations
to prescribe EPO and iron to their patients. All interviewed nephrologists were
fully aware of the guidelines, but 2 nephrologists argued that they might
withhold iron administration with ferritin levels higher than 500 ng/mL (500
µg/L), whereas 1 nephrologist would continue iron administration with ferritin
levels as high as 1,200 ng/mL (1,200 µg/L). A separate subanalysis did not show
any significant difference in EPO and iron administration pattern among
nephrologists who provided care to the study patients. Only 7 patients did not
receive any EPO during the 3 months of study observation, and the average
hematocrit level on all these 7 patients was higher than 38%.
Malnutrition inflammation
score
Using the 7 components of the conventional Subjective
Global Assessment of Nutrition (SGA), a semiquantitative scale with 3 severity
levels, and combining it with 3 new elements (body mass index, serum albumin,
and total iron binding capacity [TIBC] to represent serum transferrin) in
incremental fashion, the so-called Malnutrition-Inflammation Score (MIS) with 10
components has been created.17
Each MIS component has 4 levels of severity from 0 (normal) to 3 (very severe).
The sum of all 10 MIS components ranges from 0 to 30 denoting the increasing
degree of severity. In a recent prospective study on 83 MHD patients, the MIS
was compared with the conventional SGA and its refinements, anthropometry, near
infrared measured body fat percentage, laboratory measures including serum
C-reactive protein (CRP), and 12-month prospective hospitalization and mortality
rates.17
The MIS was found to be a comprehensive scoring system with significant
associations with prospective hospitalization and mortality as well as measures
of nutrition, inflammation, and anemia in MHD patients, and was superior to
conventional SGA and to individual laboratory values as a predictor of dialysis
outcome and an indicator of MICS.
In this study, MHD patients were
scored by 10 collaborating renal dietitians who were trained adequately for this
purpose. To evaluate the degree of reproducibility, the MIS was reassessed
randomly by a physician on a subset of 24 patients without reference to the
first MIS evaluation. The correlation coefficient (r) between the two MIS
assessments was 0.88 denoting a good degree of reproducibility.
Laboratory evaluation
Blood
samples were obtained in early October 2001 from group 1 and in early January
2002 from group 2 patients and coincided chronologically with the quarterly
routine blood tests of DaVita facilities. The laboratory values, except for
postdialysis serum urea nitrogen, which was used to calculate the urea reduction
ratio (URR), were measured immediately before the initiation of dialysis
treatment. The double-pool Kt/V was used to represent the weekly dialysis dose
and the protein equivalent of total nitrogen appearance (nPNA), also known as
normalized protein catabolic rate (nPCR) and was calculated to estimate the
daily protein intake.20
All routine laboratory measurements were performed by DaVita Laboratories
(Deland, FL) using automated methods, and the average values of each laboratory
test within the 13-week study period were calculated and used in this study.
Serum CRP and cytokines, including interleukin-6 (IL-6) and tumor
necrosis factor- (TNF-), were obtained to indicate the
presence of an inflammatory state. The high-sensitivity CRP was measured by a
turbidometric immunoassay in which a serum sample is mixed with latex beads
coated with antihuman CRP antibodies forming an insoluble aggregate (WPCI,
Osaka, Japan; mg/L; normal range, <3.0 mg/L).21,22
High-sensitivity IL-6 and TNF-
immunoassay kits based on a solid phase sandwich using recombinant human IL-6
and TNF- were used to measure
serum proinflammatory cytokines (R & D Systems, Minneapolis, MN; pg/mL;
normal range, IL-6: <9.9 pg/mL, TNF-: <4.7 pg/mL).23–25
CRP and cytokines were measured in the General Clinical Research Center Core
Laboratories of Harbor-UCLA Medical Center. Serum prealbumin was analyzed by an
antigen-antibody complex assay, and total plasma homocysteine concentrations
were determined by high-performance liquid chromatography (HPLC) at Harbor-UCLA
Clinical Laboratories.
Statistical methods
Conventional
Sudent’s t test and analysis of variance (ANOVA) were used to detect
significant differences among continuous variables between men and women and
among EPO tertiles. Chi-square method was used for categorical variables
including sex, race, ethnicity, diabetes, and route of EPO administration (IV
v SQ). To determine the significance and strength of bivariate
associations, we used Pearson´s correlation coefficient r for analyses of
associations between continuous variables. Multivariate regression analysis was
performed to obtain partial (adjusted) correlations controlled for conventional
case-mix features (sex, age, race, and presence of diabetes) and route of EPO
administration, dialysis center, ZIP code, insurance status (fully Medicaid
versus others), Kt/V (single pool), blood hematocrit, serum iron saturation
ratio, administered IV iron dose, use of angiotensin pathway blocking
medications, Charlson comorbidity score, and dialysis vintage. To calculate the
odds ratios (ORs) and their 95% confidence interval (CI) levels for highest
versus lowest EPO tertiles, we used logistic regression models after controlling
for the above-mentioned demographic variables. Therefore, the association
between variables was studied via separate multivariate models but with uniform
adjustment in each model. Fiducial limits are given as mean ± SD. Charlson score
and MIS were analyzed as continuous variables. Logarithmic conversion was
carried out for nonnormally distributed variables such as inflammatory markers.
A P value lesss than 0.05 or a 95% CI that did not span 1.0 was
considered to be statistically significant. Descriptive and multivariate
statistics were carried out with the statistical software Stata, version 7.0
(Stata Corporation, College Station, TX), and the results were verified using a
second statistical software, Statistica for Windows, Release 6.0 (Statsoft, Inc,
Tulsa, OK).
| Results | TOP |
shows some relevant descriptive data of study subjects.
 |
All MHD Patients (n = 339) | Women (n = 158) | Men (n = 181) | P Value |
| Ethnicity (Hispanics %) | 47 | 46 | 47 | 0.89 |
| Race (black %) | 30 | 34 | 28 | 0.24 |
| Diabetes (%) | 55 | 57 | 53 | 0.49 |
| SQ EPO % | 19 | 19 | 19 | 0.95 |
| History of cardiac disease (%) | 51 | 55 | 47 | 0.15 |
| ACE inhibitor or ARB (%) | 33 | 30 | 35 | 0.33 |
| Age (y) | 54.7 ± 14.5 | 56.7 ± 14.5 | 53 ± 14.3 | 0.02 |
| Dialysis vintage (mo) | 36.3 ± 33.2 | 35.4 ± 30.5 | 37.1 ± 35.5 | 0.64 |
| Post-HD weight (kg) | 72.5 ± 18.3 | 69.3 ± 17.48 | 75.4 ± 18.49 | <0.01 |
| BMI (kg/m2) | 26.7 ± 6.4 | 27.1 ± 6.1 | 26.3 ± 6.7 | 0.24 |
| Kt/V (single-pool) | 1.56 ± 0.28 | 1.66 ± 0.28 | 1.47 ± 0.25 | <0.01 |
| nPNA (nPCR) (g/kg/d) | 1.05 ± 0.22 | 1.05 ± 0.21 | 1.05 ± 0.22 | 0.82 |
| Blood hematocrit (%) | 35.4 ± 3.1 | 35.6 ± 3 | 35.2 ± 3.1 | 0.18 |
| Hemoglobin (g/dL) | 11.9 ± 1 | 11.9 ± 1 | 11.9 ± 1 | 0.84 |
| Peripheral lymphocyte count (%) | 22.3 ± 8.2 | 21.4 ± 7.4 | 23.1 ± 8.9 | 0.06 |
| Serum ferritin (ng/mL) | 635 ± 461 | 646 ± 475 | 625 ± 450 | 0.67 |
| Iron saturation ratio (%) | 32 ± 11.5 | 31.8 ± 11.7 | 32.2 ± 11.4 | 0.78 |
| TIBC (mg/dL) | 199.7 ± 37.2 | 197.2 ± 34 | 202 ± 39.6 | 0.24 |
| Albumin (g/dL) | 3.86 ± 0.33 | 3.81 ± 0.30 | 3.91 ± 0.35 | <0.01 |
| Prealbumin (mg/dL) | 28.7 ± 9.4 | 27.3 ± 8.1 | 29.9 ± 10.2 | 0.01 |
| Total cholesterol (mg/dL) | 144 ± 48.2 | 148.4 ± 48.1 | 140.2 ± 48.1 | 0.12 |
| Total homocysteine (Mg/L) | 24.6 ± 12.3 | 24.0 ± 10.2 | 25.2 ± 13.8 | 0.38 |
| CRP (mg/L) | 6.52 ± 8 | 5.95 ± 5.51 | 7.02 ± 9.65 | 0.22 |
| IL-6 (pg/mL) | 21.85 ± 57.82 | 24.74 ± 73.62 | 19.33 ± 39.2 | 0.39 |
| TNF- (pg/ml) |
8.5 ± 6.46 | 9.25 ± 8 | 7.84 ± 4.66 | 0.05 |
| LDH (u/dL) | 165.8 ± 40.8 | 171.8 ± 41 | 160.6 ± 40 | 0.01 |
| Aluminum (pg/dL) | 18 ± 8 | 17.7 ± 7.3 | 18.2 ± 8.7 | 0.65 |
| Intact PTH (mu/L) | 357 ± 380 | 371 ± 378 | 346 ± 383 | 0.55 |
| Charlson comorbidity
score* |
2.1 ± 1.5 | 2.1 ± 1.5 | 2 ± 1.6 | 0.69 |
| MIS | 6.26 ± 3.8 | 6.59 ± 3.42 | 5.97 ± 4.1 | 0.15 |
| EPO dose (U/wk) | 15.220 ± 13514 | 16,439 ± 15,016 | 14,155 ± 11,992 | 0.12 |
| EPO dose per weight (U/kg/wk) | 217 ± 187 | 244 ± 206 | 194 ± 165 | 0.01 |
| EPO responsiveness index | 1,329 ± 1,281 | 1,429 ± 1,427 | 1,242 ± 1,135 | 0.18 |
| EPO responsiveness index per weight | 3.28 ± 3.12 | 3.83 ± 3.49 | 2.8 ± 2.67 | <0.01 |
| Average IV iron dose (mg/wk) | 39.5 ± 47.52 | 38.3 ± 47.1 | 40.5 ± 48 | 0.67 |
NOTE. P values are based on Student’s t-test comparing men with women. To convert hemoglobin or albumin in g/dL to g/L, multiply by 10; ferritin in ng/mL to µg/L, multiply by 1; cholesterol in mg/dL to mmol/L, multiply by 0.02586; homocysteine is mg/L to µmol/L, multiply by 7.397. The modified Charlson comorbidity index used in this study does not include age and does not count for renal disease. P values smaller than 0.05 are bolded. |
||||
*Not adjusted for multiple comparisons. | ||||
, the study population had a heavy Hispanic contribution
of 47%, which is not an unusual ethnicity distribution in many Southern
California dialysis facilities. Thirty-percent of patients were black, and more
than half of them (55%) had diabetes. More than half (51%) of the study patients
had a history of cardiac disease. Almost one third of patients were on
angiotensin-converting enzyme (ACE) inhibitors and/or angiotension-blocking
agents that might play a role in EPO hyporesponsiveness.26
There was no statistically significant difference between these features in men
and women. Table 1 also shows descriptive analysis of continuous variables
and compares them in men and women via Student’s t test. Women were
slightly older than men and, as expected, had a lower postdialysis dry weight,
but there was no difference between men and women in terms of body mass index
(BMI). Kt/V was slightly higher in women. Men’s serum albumin and prealbumin
concentrations were 0.1 g/dL and 2.6 mg/dL higher than women’s, respectively,
but serum lactase dehydrogenase (LDH) concentration was higher in women.
Administered dose of EPO and EPO responsiveness index per kilogram of body
weight were higher in women, probably because women weighed on average 6 kg less
than men. There was no significant difference among other values between men and
women. shows correlation coefficients (r) between average EPO
dose, EPO dose per kilogram of weight, EPO responsiveness index, and relevant
nutritional and inflammatory values.
|
EPO Dose (total per week) | EPO Dose per kg Weight | EPO Responsiveness Index |
| Age | –.07 (P = .19) | –.07 (P = .18) | –.06 (P = .26) |
| Dialysis vintage | +0.5 (P = .41) | +0.6 (P = .30) | +.04 (P = .49) |
| Kt/V (single pool) | –.08 (P = .12) | +.04 (P = .48) | .09 (P = .15) |
| Iron saturation ratio | –.16 (P = .003) | –.11 (P = .04) | –.14 (P = .01) |
| Administered IV iron dose | +.16 (P = .003) | +.15 (P = .007) | +.14 (P = .01) |
| Modified Charlson comorbidity score | +.09 (P = .09) | +.07 (P = 23) | +.10 (P = .07) |
| Blood hematocrit | –.27 (P < .001) | –.29 (P < .001) | –.35 (P < .001) |
| Blood lymphocyte percentage | –.18 (P = .001)/.17 (P < .002) | –.21 (P < .001)/.21 (P < .001) | –.18 (P = .001)/.17 (P = .002) |
| MIS | +.17 (P = .003)/.13 (P = .03) | +.24 (P < .001)/.23 (P < .001) | +.17 (P = .002)/.13 (P = .02) |
| nPNA (nPCR) | –.01 (P = .83)/.03 (P = .55) | +.03 (P = .55)/.06 (P = .30) | –.03 (P = .64)/.02 (P = .65) |
| Serum albumin | –.11 (P = .04)/.05 (P = .39) | –.13 (P = .01)/.11 (P = .07) | –.14 (P = .009)/.07 (P = .24) |
| Ferritin | +.05 (P = .38)/.04 (P = .45) | +.07 (P = .21)/.04 (P = .48) | +.07 (P = .17)/.06 (P = .28) |
| TIBC | –.15 (P = .006)/.15 (P = .006) | –.21 (P < .001)/.21 (P < .001) | –.17 (P = .002)/.15 (P = .006) |
| Total cholesterol | –.19 (P < .001)/.19 (P < .001) | –.20 (P < .001)/.22 (P < .001) | –.18 (P < .001)/.18 (P < .001) |
| Prealbumin | –.19 (P < .001)/.15 (P = .006) | –.22 (P < .001)/.20 (P < .001) | –.19 (P < .001)/.15 (P = .004) |
| CRP | +.21 (P < .001)/.18 (P < .001) | +.20 (P < .001)/.19 (P < .001) | +.20 (P < .001)/.17 (P < .001) |
| Log of CRP | +.22 (P < .001)/.18 (P < .001) | +.18 (P = .001)/.15 (P = .007) | +.22 (P < .001)/.17 (P = .001) |
| IL-6 | +.19 (P = .001)/.14 (P = .007) | +.17 (P = .002)/.12 (P = .02) | +.18 (P = .001)/.14 (P < .005) |
| Log of IL-6 | +.34 (P < .001)/.31 (P < .001) | +.32 (P < .001)/.30 (P < .001) | +.34 (P < .001)/.31 (P < .001) |
| TNF- |
+.16 (P = .003)/.18 (P = .001) | +.19 (P < .001)/.20 (P < .001) | +.15 (P = .007)/.18 (P = .002) |
| Log of TNF- |
+.17 (P = .001)/.18 (P = .001) | +.20 (P < .001)/.19 (P < .001) | +.16 (P = .002)/.18 (P = .001) |
| Aluminum | –.01 (P = .91)/.13 (P = .05) | +.06 (P = 35)/.02 (P = .70) | –.01 (P = .81)/.12 (P = .05) |
| Intact PTH | +.18 (P = .001)/.11 (P = .05) | +.15 (P = .007)/.09 (P = .12) | +.17 (P = .009)/.11 (P = .05) |
| LDH | +.27 (P < .001)/.19 (P < .001) | +.30 (P < .001)/.24 (P < .001) | +.28 (P < .001)/.20 (P < .001) |
NOTE. In each cell, the first value is the unadjusted (bivariate) r with its Pearson P value in parentheses, and the second value is partial correlation based on multivariate regression analysis adjusted for case-mix (age, sex, race, diabetes), subcutaneous EPO administration, dialysis center, ZIP code, insurance status (fully Medicaid versus others), Kt/V (single pool), blood hematocrit level, serum iron saturation ratio, administered IV iron dose, use of angiotensin pathway blocking medications, Charlson comorbidity score, and dialysis vintage (partial correlation P values are in parentheses). Correlation coefficient values equal or greater than 0.20 are bolded. (P values are not adjusted for multiple comparisons.) | |||
shows scatter diagrams for serum IL-6 and total
cholesterol concentrations plotted against administered EPO dose. . Patients in the highest (third) EPO tertile had a higher
proportion of women and received subcutaneous EPO more often than the other 2
tertiles.
| Fig 1. Bivariate
correlations, in form of scatter diagrams, between administered EPO (per
kilogram body weight per week, averaged over a 3-month period) and IL-6 to
represent inflammation (A) and total serum cholesterol to represent
nutritional state (B). To convert cholesterol in mg/dL to mmol/L, multiply
by 0.02586. |
)
|
|
|
|
First (Lowest) EPO/kg Tertile | Second (Middle) EPO/kg Tertile | Third (Highest) EPO/kg Tertile | P Value |
| Sex (men %) | 63 | 54 | 43 | 0.01 |
| Ethnicity (Hispanics %) | 39 | 53 | 49 | 0.10 |
| Race (blacks %) | 26 | 29 | 36 | 0.21 |
| Diabetes % | 54 | 53 | 57 | 0.86 |
| SQ EPO (%) | 11 | 26 | 20 | 0.01 |
| EPO dose (U/wk) | 4,954 ± 3,331 | 13,209 ± 4,581 | 27,495 ± 16,020 | <0.001 |
| EPO dose per weight (U/kg/wk) | 64 ± 36 | 178 ± 37 | 408 ± 202 | <0.001 |
| EPO respons index (EPO/Hb) | 415 ± 286 | 1,117 ± 418 | 2,456 ± 1,590 | <0.001 |
| EPO respons./wt (EPO/Hb/kg) | 0.88 ± 0.51 | 2.55 ± 0.84 | 6.41 ± 3.49 | <0.001 |
| MIS | 5.8 ± 3.4 | 5.6 ± 3.8 | 7.5 ± 4.0 | <0.001 |
| Blood Hematocrit (%) | 36.0 ± 2.7 | 35.6 ± 2.9 | 34.6 ± 3.4 | 0.001 |
| periph. lymphocyte count (%) | 23.8 ± 8.1 | 23.3 ± 8.7 | 19.7 ± 7.3 | <0.001 |
| Serum ferritin (ng/mL) | 625 ± 378 | 600 ± 480 | 678 ± 515 | 0.43 |
| Iron saturation ratio (%) | 33.8 ± 10.9 | 32.5 ± 11.4 | 29.9 ± 11.9 | 0.03 |
| TIBC (mg/dL) | 205.4 ± 36.9 | 202.8 ± 36.3 | 191.1 ± 37.0 | 0.008 |
| Albumin (g/dL) | 3.88 ± 0.32 | 3.89 ± 0.32 | 3.81 ± 0.34 | 0.14 |
| Prealbumin (mg/dL) | 30.7 ± 9.8 | 28.8 ± 8.8 | 26.5 ± 9.1 | 0.003 |
| Cholesterol (mg/dL) | 152.3 ± 43.4 | 143.5 ± 51.7 | 135.9 ± 48.2 | 0.04 |
| CRP (ng/mL) | 6.26 ± 8.18 | 5.31 ± 4.92 | 8.00 ± 9.91 | 0.03 (0.04*) |
| IL-6 (ng/mL) | 12.26 ± 19.09 | 23.64 ± 75.73 | 29.66 ± 61.92 | 0.07 (<0.001*) |
| TNF- (ng/mL) |
7.63 ± 6.78 | 8.77 ± 6.15 | 9.09 ± 6.41 | 0.20 (0.08*) |
| LDH (mg/dL) | 153.6 ± 32.3 | 159.3 ± 31.5 | 184.6 ± 49.4 | <0.001 |
| Aluminum (ng/mL) | 18.0 ± 8.0 | 16.9 ± 6.1 | 19.2 ± 9.8 | 0.17 |
| Intact PTH (mcu/L) | 337 ± 354 | 350 ± 345 | 385 ± 437 | 0.61 |
| IV iron dose (mg/mo) | 25.4 ± 31.2 | 45.7 ± 48.3 | 47.4 ± 56.6 | <0.001 |
| BMI (kg/m2) | 27.7 ± 7.6 | 27.0 ± 5.8 | 25.3 ± 5.5 | 0.01 |
| Charlson comorbidity score | 1.9 ± 1.5 | 2.0 ± 1.6 | 2.2 ± 1.5 | 0.31 |
| Age (y) | 55.2 ± 14.6 | 53.3 ± 14.2 | 55.6 ± 14.6 | 0.42 |
| Dialysis vintage (mo) | 36.6 ± 32.6 | 33.6 ± 32.6 | 38.7 ± 34.6 | 0.52 |
NOTE. To convert ferritin in ng/mL to µg/L, multiply by 1; albumin in g/dL to g/L, multiply by 10; cholesterol in mg/dL to mmol/L, multiply by 0.02584. The listed P values are based on ANOVA. |
||||
*For inflammatory markers (CRP, IL-6, and TNF- ), additional
P values based on logarithmic transformation of these markers have
been listed in parentheses. | ||||
,
which did not achieve statistical significance despite a similar trend.
Analogous association was observed with serum LDH level as well as required EPO
dose. Figure 2 depicts bar diagrams of serum IL-6 and total cholesterol
concentrations across EPO tertiles. shows logistic regression estimated ORs and 95% CIs for
highest versus lowest EPO tertile adjusted for the same covariates as in Table
2.
| Fig 2. Serum levels of IL-6
(A) and total cholesterol concentration (B) within the tertiles of
administered EPO (per kilogramof body weight per week, averaged over a
3-month period). To convert cholesterol in mg/dL to mmol/L, multiply by
0.02586. |
)
|
|
|
|
OR | 95% CI | P value |
MIS (for each 5 unit
 ) |
1.64 | 1.02–2.65 | 0.04 |
| Serum CRP (for each 10
mg/L ) |
1.25 | 0.88–1.78 | 0.21 |
| Log of CRP (for each 1
unit ) |
1.44 | 0.75–2.81 | 0.27 |
| IL-6 (for each 10 pg/mL
) |
1.20 | 1.02–1.42 | 0.03 |
| Log of IL-6 (for each 1
unit ) |
4.45 | 1.94–10.20 | <0.001 |
| TNF- (pg/mL) |
1.66 | 1.03–2.69 | 0.04 |
| Log of TNF- (for each 1 unit ) |
4.91 | 1.05–22.82 | 0.04 |
Albumin (for each 0.5
g/dL  ) |
1.12 | 0.65–1.94 | 0.67 |
| Prealbumin (for each 10
mg/dL ) |
1.65 | 1.13–2.42 | 0.01 |
| TIBC (for each 50 mg/dL
) |
2.15 | 1.30–3.56 | 0.003 |
| Cholesterol (for each
50 mg/dL ) |
2.22 | 1.49–3.30 | <0.001 |
| LDH (for each 50 mg/dL
) |
3.06 | 1.75–5.38 | <0.001 |
| Ferritin (for each 500
ng/mL ) |
1.42 | 0.95–2.13 | 0.09 |
| Blood lymphocyte count
(for each 10 % ) |
2.74 | 1.56–4.82 | <0.001 |
NOTE. OR and 95% CI for tertiles of “weekly administered EPO dose per body weight” comparing the odds of highest versus lowest EPO dose tertile, adjusted for case-mix (age, sex, race, DM), subcutaneous EPO administration, dialysis center, ZIP code, insurance status (fully Medicaid versus others), Kt/V (single pool), blood hematocrit, serum iron saturation ratio, administered IV iron dose, use of angiotensin pathway blocking medications, Charlson comorbidity score, and dialysis vintage by means of logistic regression analyses. The magnitude and direction of changes for each variable to calculate OR are based on considering the reported standard deviations (see Table 1 ) and after being rounded to practical
increments or decrements for clinical use. | |||
) and after being rounded to practical increments or
decrements that appear to be relevant for clinical use. Associations and their
directions were essentially similar to those seen in Tables 2 and 3. Each 5-unit increase in MIS increased the risk of EPO
hyporesponsiveness by 64%. Similar associations also were observed between
inflammatory markers and refractory anemia. Once again, most nutritional markers
including serum concentrations of prealbumin, TIBC, and total cholesterol as
well as peripheral lymphocyte count had a negative association with EPO dose in
that their decrease led to higher risk of EPO hyporesponsiveness. | Discussion | TOP |
20 mg/L was
80% higher than in patients with CRP less than 20 mg/L. Goicoechea et al28
reported a significant direct correlation between IL-6 and TNF- production and weekly EPO dose per
kilogram of body weight in 34 MHD patients. Gunnell et al6
examined 92 MHD and 36 peritoneal dialysis patients and found that the
acute-phase response in form of high-serum CRP, high ferritin, and low
transferrin concentrations were predictors of EPO resistance. Sitter et
al29
found an association between dialysis-induced rise in serum IL-6 level and
increasing required EPO dosage in 30 MHD patients. In our current study, the
logarithm of serum IL-6 level had the strongest correlation with the required
EPO dose, and the association remained statistically significant in different
statistical analyses and after multivariate adjustments. Both serum CRP and
TNF- also showed a similar trend,
and their associations with EPO dose remained significant in some but not all
analyses we conducted in this study. have been shown to
inhibit EPO production in vitro.35
Furthermore, increased release or activation of inflammatory cytokines, such as
IL-6 or TNF-, has been shown to
have suppressive effect on erythropoiesis.36
IL-6 and IL-1 have been found to antagonize EPO’s ability to stimulate bone
marrow proliferation in culture.37
However, other studies did not find such effects or found paradoxical
associations.38
Finally, patients with inflammation may be more prone to gastrointestinal
bleeding.12,32