Sei sulla pagina 1di 4

Journal of the American Society of Hypertension 6(6) (2012) 439442

Debate from the American Society of Hypertension Annual Meeting

Debate from the 2012 ASH Annual Scientific Sessions: should blood
pressure be reduced in hemodialysis patients? pro position
Rajiv L. Agarwal, MD
Indiana University School of Medicine and Richard L. Roudebush Veterans Administration Medical Center, Indianapolis, IN
Manuscript received August 16, 2012 and accepted August 26, 2012


Among chronic hemodialysis patients with hypertension, blood pressure should be lowered. Blood pressure reduction with
antihypertensive drugs does not increase mortality; in contrast, meta-analysis of randomized trials suggests that treatment
of hypertension in this high-risk population may, in fact, improve cardiovascular outcomes. The association of low blood
pressure with increased mortality in longitudinal studies should not be considered as evidence against lowering blood pres-
sure. Lowering blood pressure among hypertensive patients should primarily be done by sodium restriction and dry-weight
reduction. Treatment is perhaps better directed to home blood pressure than pre- or post-dialysis blood pressure recordings.
Although no firm data are available, it appears that treating home blood pressure to <140/90 mm Hg appears reasonable.
Nonetheless, all blood pressure recordings during dialysis are important to ensure patient safety. Adequately designed
and powered randomized trials are needed to examine the notion that blood pressure lowering and, if so, to what level of
blood pressure will improve clinically meaningful outcomes among chronic dialysis patients. J Am Soc Hypertens
2012;6(6):439442. Published by Elsevier Inc on behalf of American Society of Hypertension.

Whether blood pressure should be lowered among dialysis a risk ratio of 2.62 compared with a reference blood pressure
patients is a matter of controversy because epidemiological of 140149 mm Hg.2 In contrast, noncardiac-related cause
studies nearly uniformly suggest that lower blood pressure of death was associated with a risk ratio of 1.55 when the
among hemodialysis patients is associated with greater systolic blood pressure was <110 mm Hg compared with
mortality.1 The first study in the United States that convinc- a reference range of 140149 mm Hg.2
ingly demonstrated that lower blood pressure is associated The next study, published in 1999, the United States Renal
with an increased mortality was performed in a cohort of Data System case mix adequacy study, demonstrated that
patients dialyzed at facilities managed by a nonprofit dialysis a pre-dialysis blood pressure of <109 mm Hg systolic was
provider.2 In this observational study involving several thou- associated with a risk ratio of cardiovascular death of 1.86
sand patients, Zager et al demonstrated that the relative death and a blood pressure of 110119 mm Hg was associated
rate was not changed with increasing levels of blood pressure with a risk ratio of 1.27.3 The reference pre-dialysis systolic
measured post-dialysis.2 However, if post-dialysis blood pres- blood pressure (with risk ratio of 1) was 120 to 149 mm
sure was modeled as a time-dependent covariate, then it was Hg.3 In 2005, Kalantar-Zadeh et al confirmed the inverse rela-
found that a lower blood pressure was associated with tionship of cardiovascular death and blood pressure in the
increased mortality.2 A post-dialysis systolic blood pressure hemodialysis population, which they described as the reverse
of <110 mm Hg was associated with a risk ratio of 2.04 for epidemiology of hypertension.4 Using the database of a dial-
death compared with a reference blood pressure post dialysis ysis provider (DaVita), the authors demonstrated a J-shaped
of 140149 mm Hg.2 Cause-specific death rates showed an relationship of cardiovascular death even after adjustments
even more remarkable relationship. For instance, post- for several risk factors.4 Cardiovascular death was similarly
dialysis blood pressure of <110 mm Hg was associated with strongly and inversely associated with post-dialysis systolic
blood pressure.4 Post-dialysis systolic blood pressure of
Funding: NIH 2R01-DK062030-08.
<120 mm Hg was associated with an increased mortality.
Corresponding author: Rajiv L. Agarwal, MD, Professor of The following year, Li et al confirmed these observations in
Medicine, Indiana University and VAMC, 1481 West 10th Street, yet another analysis of a dialysis provider (Fresenius).5 Pre-
Indianapolis, IN 46202. Tel: (317) 988-2241. dialysis systolic blood pressure <120 mm Hg was associated
E-mail: with a hazard ratio of 3.5 (reference predialysis systolic blood
1933-1711/$ - see front matter Published by Elsevier Inc on behalf of American Society of Hypertension.
440 R.L. Agarwal / Journal of the American Society of Hypertension 6(6) (2012) 439442

pressure <180 mm Hg) for all-cause mortality.5 The time- hemodialysis patients suggest a benefit of blood pressure
varying model for death yielded estimates of hazards of 67.5 lowering. In a meta-analysis of eight trials, the overall reduc-
These observations can perhaps be explained by the high tion in cardiovascular events was 36% (P < .05).8 In further
comorbidity that exists in dialysis patients. For example, in examination of these studies, it was noted that when normo-
a low-risk population (such as those with uncomplicated tensive patients were included, the risk reduction for adverse
hypertension, absence of diabetes, and heart failure), a direct cardiovascular events was only 12% in contrast to a 55% risk
relationship would be anticipated between increasing blood reduction when hypertensive patients were treated with anti-
pressure and mortality.6 However, in a high-risk population hypertensive agents.9
(such as the patients who have presence of heart failure and To date, no randomized trial with hard outcomes in dialysis
systemic inflammation), this relationship could be hypothe- patients has focused on blood pressures measured outside the
sized to be inverted. This occurs because blood pressures in dialysis unit. It is important to recognize that dialysis unit
these patients the severity of the underlying illness. blood pressures are poor estimates of inter-dialytic ambula-
Compared with those with essential hypertension, dialysis tory blood pressure.10 For example, a meta-analysis demon-
patients have a much higher proportion of patients with strated that the pre-dialysis systolic blood pressure
the heart failure and systemic inflammation; therefore, overestimated ambulatory blood pressure by 16.7 mm Hg
a U-shaped relationship could be anticipated.6 and post-dialysis values underestimated by 1.6 mm Hg.11
These observational, cohort studies did not lower blood However, the standard deviations of the estimated difference
pressure; in fact, none of these studies had any information between dialysis unit blood pressure recordings and ambula-
on antihypertensive medications. The fall in blood pressure tory blood pressure measurements in individual studies are
was spontaneous and may have been related to disease wide. Accordingly, in individual patients, post-dialysis or
states. Accordingly, one has to be cautious in extrapolating pre-dialysis systolic blood pressures cannot reliably predict
these studies. Cohort studies examine association but not ambulatory blood pressure.11 In contrast, home blood pres-
causality. sure greater than 150 mm Hg has a sensitivity of 80% and
The distinction between achieving a lower blood pressure specificity of 84% in diagnosing interdialytic ambulatory
and targeting a lower blood pressure can be demonstrated hypertension.12 Furthermore, systolic home blood pressure
through a clever analysis of the African American Study is similar to ambulatory blood pressure in predicting echocar-
of Kidney disease. In that study, target mean arterial pressure diographic left ventricular hypertrophy.13 This is not so for
for the two randomized groups was <92 mm Hg or dialysis unit blood pressurerecorded values. In fact, out-
<115 mm Hg.7 The primary end point was the rate of decline of-dialysis unit blood pressures are of greater prognostic
in the measured glomerular filtration rate (GFR), which was value compared with those obtained within the dialysis
not different over 4 years between the two groups. The unit.14 In a hemodialysis cohort of 150 patients followed
composite end point of 50% decline in GFR, a decline of at for 2 years, 31% of the patients died.14 Quartiles of systolic
least 25 mL/min/1.73 m2 in GFR, end-stage renal disease, blood pressure were associated with mortality in a Cox
or death was also no different between the randomized model. For the pre-dialysis or post-dialysis blood pressure,
groups.7 This leads to the conclusion that targeting blood no statistical significance was seen. However, home blood
pressure more aggressively does not reduce the risk of pressure of 125145 mm Hg was associated with the best
progression of kidney disease for hard outcomes in such pop- prognosis, as was an ambulatory blood pressure of
ulations.7 Achieved blood pressure analysis, however, 115125 mm Hg.14 Both home blood pressure and ambula-
yielded results that were quite different. For example, the tory blood pressure were statistically significant for predict-
mean GFR slope was higher in patients who had a mean ing prognosis.14 In a larger cohort of more than 320 patients,
follow-up mean arterial pressure of 113 mm Hg as compared pre- and post-dialysis blood pressures were not found to be of
with those with a mean achieved pressure of 86 mm Hg.7 The prognostic significance for all-cause mortality.15 In contrast,
hazard ratio of the primary end point was also increased in ambulatory blood pressure quartiles were highly associated
those patients who were in the highest quartile of mean with death, as were home blood pressure recordings.15
>113 mm Hg. Accordingly, analyses based on achieved Finally, in a trial designed to treat hypertension using home
blood pressure led to markedly different inferences than the blood pressure versus dialysis unit blood pressure, it was
intention-to-treat analysis.7 This may be attributed to con- noted that over a 6-month period in patients randomized to
founding of achieved blood pressure by comorbidities, dialysis unit obtained blood pressure treatment, systolic in-
disease severity, or nonadherence.7 As a result, the authors terdialytic ambulatory blood pressure changed from 149 to
concluded that clinicians and policy makers should exercise 150 mm Hg (P NS).16 In those patients in whom treatment
caution when making treatment recommendations based on was guided using home blood pressure, blood pressure
analyses relating outcomes to achieved blood pressures.7 changed from 149 to 142, which was statistically signifi-
Although adequately powered randomized trials among cant.16 Although the study failed to find improvement in
dialysis patients are not yet available, meta-analyses of left ventricular hypertrophy, it was underpowered to detect
randomized controlled trials of antihypertensive drugs in this difference. Given the various benefits of home blood
R.L. Agarwal / Journal of the American Society of Hypertension 6(6) (2012) 439442 441

pressure monitoring emphasized in the United States17,18 and mortality in longitudinal studies should not be considered
European guidelines,19,20 it appears that this should be the as evidence against lowering blood pressure. Blood pressure
basis for the treatment of hypertension for most dialysis reduction among hypertensive patients should primarily be
patients. accomplished by sodium restriction and dry-weight reduc-
How to get the blood pressure down has been an important tion. Treatment is perhaps best directed to home blood
conundrum among nephrologists. The first-line management pressure rather than pre- or post-dialysis blood pressure
should be through volume reduction. In fact, Dr. Belding recordings. Nonetheless, all blood pressures measured
Scribner, a pioneer in dialysis treatment in the United States during dialysis are important to ensure patient safety. Finally,
remarked that In the case of dialysis patients, the slow adequately designed and well-powered randomized trials are
normal level of extracellular volume is maintained by needed to examine the notion whether blood pressure
a powerful tool, ultrafiltration, which is, if properly used lowering will improve clinically meaningful outcomes
along with dietary sodium restriction, the only proven among chronic dialysis patients.
method of controlling blood pressure in the hemodialysis
populations. In my view, the answer to the question can anti-
hypertensive medications control blood pressure in hemodi- References
alysis patients is no.21 Therefore additional volume
reduction through probing dry weight can serve as an impor- 1. Agarwal R. Hypertension and survival in chronic hemo-
tant tool to control hypertension in these patients. In the only dialysis patientspast lessons and future opportunities.
randomized trial that has examined this issuethe Dry Kidney Int 2005;67:113.
weight Reduction In hypertensive hemodialysis Patients 2. Zager PG, Nikolic J, Brown RH, Campbell MA,
(DRIP) randomized trialit was found that interdialytic Hunt WC, Peterson D, et al. U curve association of
ambulatory blood pressure was reduced 6.9 mm Hg systolic blood pressure and mortality in hemodialysis patients.
within 4 weeks and 3.1 mm diastolic in the same period.22 Medical Directors of Dialysis Clinic, Inc [published
This was not associated with deterioration in quality of life erratum appears in Kidney Int 1998 Oct;54(4):1417].
but may have been associated with a slight increase in Kidney Int 1998;54:5619.
adverse events that were not statistically significant.22 3. Port FK, Hulbert-Shearon TE, Wolfe RA,
Clearly, further studies are needed to examine this notion Bloembergen WE, Golper TA, Agodoa, et al. Predialy-
further, but for now it appears that volume reduction should sis blood pressure and mortality risk in a national
be attempted as first-line therapy to control hypertension in sample of maintenance hemodialysis patients. Am J
most chronic hemodialysis patients. Kidney Dis 1999;33:50717.
Because few techniques are available to assess volemia, 4. Kalantar-Zadeh K, Kilpatrick RD, McAllister CJ,
relative plasma volume slopes during the dialysis procedure Greenland S, Kopple JD. Reverse epidemiology of
measured through commercially available equipment can be hypertension and cardiovascular death in the hemodial-
used along with the overall clinical scenario to judge volume ysis population: the 58th annual fall conference and
status.23 In fact, in the DRIP trial, participants who had dry scientific sessions. Hypertension 2005;45:8117.
weight probed had steepening of relative plasma volume 5. Li Z, Lacson E Jr, Lowrie EG, Ofsthun NJ,
slopes.24 Furthermore, those with the flattest slopes at base- Kuhlmann MK, Lazarus JM, et al. The epidemiology
line experienced the greatest reduction in blood pressure. of systolic blood pressure and death risk in hemodial-
In contrast, on probing dry weight the bottom quartile had ysis patients. Am J Kidney Dis 2006;48:60615.
no reduction in blood pressure. Preliminary data suggest 6. Agarwal R. Exploring the paradoxical relationship of
that in comparison to flatter slopes, steeper relative plasma hypertension with mortality in chronic hemodialysis.
volume slopes are associated with lower mortality.25 This Hemodialysis Int 2004;8:20713.
mortality benefit persists after multifactorial adjustments. 7. Davis EM, Appel LJ, Wang X, Greene T, Astor BC,
Taken together, these data suggest that lowering volume Rahman M, et al. Limitations of analyses based on
may be an effective strategy to improve cardiovascular achieved blood pressure: lessons from the African
outcomes in these patients. American study of kidney disease and hypertension
In summary, the current evidence suggests that among trial. Hypertension 2011;57:10618.
chronic hemodialysis patients with hypertension, blood pres- 8. Heerspink HJ, Ninomiya T, Zoungas S, de ZD,
sure should be lowered. Although no firm data are available, Grobbee DE, Jardine MJ, et al. Effect of lowering blood
it appears that treating home blood pressure to <140/90 mm pressure on cardiovascular events and mortality in
Hg appears reasonable. Blood pressure lowering does not patients on dialysis: a systematic review and meta-
increase mortality; in contrast, meta-analysis of randomized analysis of randomised controlled trials. Lancet 2009;
trials suggests that treatment of hypertension in this high-risk 373:100915.
population may in fact improve cardiovascular outcomes. 9. Agarwal R, Sinha AD. Cardiovascular protection with
The association of low blood pressure with increased antihypertensive drugs in dialysis patients: systematic
442 R.L. Agarwal / Journal of the American Society of Hypertension 6(6) (2012) 439442

review and meta-analysis. Hypertension 2009;53: 18. Pickering TG, Miller NH, Ogedegbe G, Krakoff LR,
8606. Artinian NT, Goff D. Call to action on use and reim-
10. Thompson AM, Pickering TG. The role of ambulatory bursement for home blood pressure monitoring: a joint
blood pressure monitoring in chronic and end-stage scientific statement from the American Heart Associa-
renal disease. Kidney Int 2006;70:10007. tion, American Society Of Hypertension, and Preven-
11. Agarwal R, Peixoto AJ, Santos SF, Zoccali C. Pre and tive Cardiovascular Nurses Association. Hypertension
post dialysis blood pressures are imprecise estimates of 2008;52:1029.
interdialytic ambulatory blood pressure. Clin J Am Soc 19. OBrien E, Asmar R, Beilin L, Imai Y, Mallion JM,
Nephrol 2006;1:38998. Mancia G, et al. European Society of Hypertension
12. Agarwal R, Andersen MJ, Bishu K, Saha C. Home recommendations for conventional, ambulatory and
blood pressure monitoring improves the diagnosis of home blood pressure measurement. J Hypertens 2003;
hypertension in hemodialysis patients. Kidney Int 21:82148.
2006;69:9006. 20. Parati G, Stergiou GS, Asmar R, Bilo G, de LP, Imai Y,
13. Agarwal R, Brim NJ, Mahenthiran J, Andersen MJ, et al. European Society of Hypertension guidelines for
Saha C. Out-of-hemodialysis-unit blood pressure is blood pressure monitoring at home: a summary report
a superior determinant of left ventricular hypertrophy. of the Second International Consensus Conference on
Hypertension 2006;47:628. Home Blood Pressure Monitoring. J Hypertens 2008;
14. Alborzi P, Patel N, Agarwal R. Home blood pressures 26:150526.
are of greater prognostic value than hemodialysis unit 21. Scribner BH. Can antihypertensive medications control
recordings. Clin J Am Soc Nephrol 2007;2:122834. BP in haemodialysis patients: yes or no? Nephrol Dial
15. Agarwal R. Blood pressure and mortality among hemo- Transplant 1999;14:2599601.
dialysis patients. Hypertension 2010;55:7628. 22. Agarwal R, Alborzi P, Satyan S, Light RP. Dry-weight
16. da Silva GV, de Barros S, Abensur H, Ortega KC, reduction in hypertensive hemodialysis patients
Mion D Jr. Home blood pressure monitoring in blood (DRIP): a randomized, controlled trial. Hypertension
pressure control among haemodialysis patients: an 2009;53:5007.
open randomized clinical trial. Nephrol Dialysis Trans- 23. Sinha AD, Agarwal R. Can chronic volume overload be
plant 2009;24:380511. recognized and prevented in hemodialysis patients?
17. Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, The pitfalls of the clinical examination in assessing
Hill MN, et al. Recommendations for blood pressure volume status. Semin Dial 2009;22:4802.
measurement in humans and experimental animals: 24. Sinha AD, Light RP, Agarwal R. Relative plasma volume
part 1: blood pressure measurement in humans: a state- monitoring during hemodialysis aids the assessment of
ment for professionals from the Subcommittee of dry weight. Hypertension 2010;55:30511.
Professional and Public Education of the American 25. Agarwal R. Hypervolemia is associated with increased
Heart Association Council on High Blood Pressure mortality among hemodialysis patients. Hypertension
Research. Hypertension 2005;45:14261. 2010;56:5127.