co m m e nta r y
Hypercalciuria is a feature of patients
carrying the WNK4 Q562E mutation and
is not present in patients carrying WNK1
mutations. As disease-causing mutations
of WNK4 enhance TRPV5 to an extent
comparable to that of wild-type WNK4
(J-B Peng et al., J Am Soc Nephrol 2004;
15: 62A, abstr.), it may be that other fac-
tors come into play in the hypercalciuria
seen in PHA II. Distal tubular calcium
reabsorption is inversely related to sodium
reabsorption, and therefore the increase
in calcium excretion may be linked to the
elevation in distal tubular sodium reab-
sorption. In addition, the transcellular
epithelial calcium channels TRPV5 and
TRPV6 are also inhibited by acidosis.
Finally, WNK1 is also a regulator of the
ion transporters ENaC and ROMK, which
may provide a basis for the role of WNK1
in the molecular pathogenesis of PHA II.
New lines of evidence indicate that WNK1
and WNK4 are upstream of two STE20-
type kinases, SPAK and OSR1, that regulate
members of the SLC12A family of elec-
troneutral cation-Cl– cotransporters.10,11
Thus, identifying at a cellular and molecu-
lar level the differences between wild-type
WNK4 and disease-causing mutants and
their effects on transport protein activity
and expression levels should provide fur-
ther insight into the mechanisms underly-
ing the pathogenesis of PHA II.
REFERENCES
1. Xu B, English JM, Wilsbacher JL et al. WNK1, a novel
mammalian serine/threonine protein kinase
lacking the catalytic lysine in subdomain II. J Biol
Chem 2000; 275: 16795–16801.
2. Wilson FH, Disse-Nicodeme S, Choate KA et al.
Human hypertension caused by mutations in WNK
kinases. Science 2001; 293: 1107–1112.
3. Mayan H, Vered I, Mouallem M et al.
Pseudohypoaldosteronism type II: marked
sensitivity to thiazides, hypercalciuria,
normomagnesemia, and low bone mineral
density. J Clin Endocrinol Metab 2002; 87: 3248–
3254.
4. Yang CL, Angell J, Mitchell R, Ellison DH. WNK
kinases regulate thiazide-sensitive Na-Cl
cotransport. J Clin Invest 2003; 111: 1039–1045.
5. Wilson FH, Kahle KT, Sabath E et al. Molecular
pathogenesis of inherited hypertension with
hyperkalemia: the Na-Cl cotransporter is inhibited
by wild-type but not mutant WNK4. Proc Natl Acad
Sci USA 2003; 100: 680–684.
6. Kahle KT, Wilson FH, Leng Q et al. WNK4 regulates
the balance between renal NaCl reabsorption and
K + secretion. Nat Genet 2003; 35: 372–376.
7. Fu Y, Subramanya A, Rozansky D, Cohen DM. WNK
kinases influence TRPV4 channel function and
localization. Am J Physiol Renal Physiol advance
online publication, January 10 2006, doi:10.1152/
ajprenal.00391.2005.
2118
8. Yamauchi K, Rai T, Kobayashi K et al. Disease- modulated by the interaction of two kinases:
causing mutant WNK4 increases paracellular Ste20-related proline-alanine-rich kinase and
chloride permeability and phosphorylates WNK4. Am J Physiol Cell Physiol 2006; 290: C134–
claudins. Proc Natl Acad Sci USA 2004; 101: C142.
4690–4694. 12. Yang CL, Zhu X, Wang Z et al. Mechanisms of
9. Kahle KT, MacGregor GG, Wilson FH et al. WNK1 and WNK4 interaction in the regulation of
Paracellular Cl– permeability is regulated by thiazide-sensitive NaCl cotransport. J Clin Invest
WNK4 kinase: insight into normal physiology and 2005; 115: 1379–1387.
hypertension. Proc Natl Acad Sci USA 2004; 101: 13. Cai H, Cebotaru V, Wang YH et al. WNK4 kinase
14877–14882. regulates surface expression of the human
10. Vitari AC, Deak M, Morrice NA, Alessi DR. The sodium chloride cotransporter in mammalian
WNK1 and WNK4 protein kinases that are cells. Kidney Int 2006; 69: 2162–2170.
mutated in Gordon’s hypertension syndrome 14. Schambelan M, Sebastian A, Rector FC Jr.
phosphorylate and activate SPAK and OSR1 Mineralocorticoid-resistant renal hyperkalemia
protein kinases. Biochem J 2005; 391: 17–24. without salt wasting (type II pseudohypoald
11. Gagnon KB, England R, Delpire E. Volume osteronism): role of increased renal chloride
sensitivity of cation-Cl– cotransporters is reabsorption. Kidney Int 1981; 19: 716–727.
see original article on page 2155
Chronic kidney disease in the
elderly: is it really a premise for
overwhelming renal failure?
F Locatelli1 and P Pozzoni1
Increasingly, the majority of patients being diagnosed as affected by
chronic kidney disease (CKD) are elderly. Nonetheless, only a rather
small proportion of elderly CKD patients actually progress to end-stage
renal disease, whereas many more die before this stage is reached,
largely because of cardiovascular disease. This underscores the urgent
need for cardiovascular prevention, even more importantly than for
renoprotection, among elderly patients with CKD.
Kidney International (2006) 69, 2118–2120. doi:10.1038/sj.ki.5001547
Chronic kidney disease (CKD) is becom- first evidence of serum creatinine levels
ing increasingly common. The United greater than 2 mg per dl in the course of
States Renal Data System has reliably 1 year): the incidence rate among patients
estimated that the number of patients on aged more than 75 years was almost seven
maintenance dialysis in the United States times that of patients aged 20–39 years
will double over the next few years,1 and a (619 versus 92 new cases per million
relatively large number of newly diagnosed population) and more than twice that of
CKD patients every year (incident CKD patients aged 40–59 years (619 versus 264
patients) are elderly. In Europe, an epide- new cases per million population).2 The
miological survey of the Île-de-France area increased incidence of CKD among the
showed a striking age-related increase in elderly translates into a similarly increased
the annual incidence of CKD (that is, the prevalence: the Third National Health and
Nutrition Examination Survey (NHANES
1Department of Nephrology and Dialysis,
III) of a nationally representative sample of
A. Manzoni Hospital, Lecco, Italy
adults in the United States between 1988
Correspondence: F Locatelli, Department of
Nephrology and Dialysis, A. Manzoni Hospital, Via and 1994 found that 7.6% of the individu-
Dell’Eremo 9/11. 23900 Lecco, Italy. als aged 60–69 years, and 25.9% of those
E-mail: nefrologia@ospedale.lecco.it aged at least 75 years, had a glomerular
Kidney International (2006) 69

filtration rate (GFR) of 15–60 ml per
minute per 1.73 m2, as against only 1.8% of
those aged 40–59 years and 0.2% of those
aged less than 40 years.3
At the same time, and in the face of this
tsunami-like wave of CKD affecting partic-
ularly older subjects, there is growing and
apparently paradoxical evidence that there
are considerably fewer patients on chronic
dialysis than patients in the earlier stages
of CKD. The NHANES III findings make it
possible to estimate that 4.7% of the United
States population (about 8.3 million indi-
viduals) has GFR levels less than 60 ml
per minute per 1.73 m2, which becomes
11.0% (19.5 million) when we also con-
sider people with slightly decreased kidney
function (GFR 60–90 ml per minute per
1.73 m2, CKD stage 2) and those with only
persistent albuminuria (CKD stage 1);3
these numbers greatly exceed the 102,000
who started chronic dialysis in the United
States in 2003, and even the 537,000 prev-
alent dialysis patients treated during the
same year.4 The main explanation for this
discrepancy between the ‘social’ burden of
CKD and that of renal replacement treat-
ments is that the number of CKD patients
progressing toward a degree of renal failure
that requires replacement treatment for
survival is much lower than the number
of patients who die before they reach
the point of end-stage renal disease, as is
clearly shown by Keith et al.5 and Foley et
al.,6 and this is largely due to the effects
of cardiovascular disease. These observa-
tions further underline what is already
known to be the devastating impact of
cardiovascular disease on the long-term
clinical outcomes of CKD patients and
stress the urgent need for strategies aimed
at preventing not only the progression of
CKD per se, but also the development of
cardiovascular complications.
In this issue, Hemmelgarn et al.7 report
that, even in a very large community-based
population, only a rather small propor-
tion (0.9%) of elderly CKD patients actu-
ally progressed to end-stage renal disease,
whereas three times as many died during
the same period. Such a relatively high
mortality rate is of course not surprising,
as the mean age of the study population
was more than 75 years, but it is worth
noting that only a minority (about 20%)
of even subjects with a baseline GFR of less
Kidney International (2006) 69
than 30 ml per minute per 1.73 m2 began
dialysis during the study. The patients with
the lowest GFR at baseline were probably
the most progressive, and so it is not sur-
prising that they showed the highest rate of
decline in renal function over time. Unfor-
tunately, the median follow-up of 2 years
was probably too short to allow a com-
plete analysis of the longitudinal changes
in renal function, which tends to limit the
applicability of the results to the whole
population of elderly CKD patients.
Unfortunately, in the study by Hem-
melgarn et al.7 there is no information
concerning the underlying nephropathy
or other variables, such as blood pressure
and proteinuria levels, which are known to
be the most important factors affecting the
progression of CKD. Above all, although it
was stratified by diabetes mellitus, gender,
and baseline renal function, the analysis of
the changes in GFR over time did not con-
sider the drugs the patients were taking (in
particular, angiotensin-converting enzyme
inhibitors and angiotensin receptor block-
ers) or the degree of blood pressure con-
trol reached with and without them. The
only information provided on this point is
that drugs blocking the renin–angiotensin
system were taken by increasing propor-
tions of the patients with decreasing base-
line renal function, which means that the
patients who progressed most during the
study were significantly more treated with
the drugs that should have protected them
against progression. It is likely that this
association can be explained by selection
bias, because it is realistic to presume that
the patients showing a faster rate of pro-
gression were more likely to be managed
by well-known nephroprotective agents;
but although only one-third of the patients
with a baseline GFR of more than 60 ml
per minute per 1.73 m2 were taking angio-
tensin-converting enzyme inhibitors and/
or angiotensin receptor blockers, their rate
of decline in renal function over the dura-
tion of the study was very slow (a mean of
0.6 ml per minute per 1.73 m2 per year).
This last observation raises the question
of whether we should reconsider the role
of renin–angiotensin system blockade as a
means of protecting against the progression
of CKD, at least in contexts such as elderly
populations in which the burden of overt
proteinuric nephropathies is believed to be
co m m e nta r y
lower than in other populations and there
is a greater potential risk of further deterio-
ration of renal function because of the high
prevalence of renal stenotic atherosclerotic
lesions and very frequent concomitant use
of diuretics and nonsteroidal anti-inflam-
matory drugs. It can likewise be argued
that epidemiological surveys should be
analyzed more cautiously, as it has been
perhaps too often claimed that the recent
reduction in the number of patients reach-
ing end-stage renal disease is mainly due
to the increasing administration of drugs
capable of counteracting progressive kid-
ney damage.
It is also particularly interesting to
note that only a relatively small propor-
tion of the CKD population enrolled in
the study made use of drugs that have
the documented potential to prevent the
development of cardiovascular events
(angiotensin-converting enzyme inhibi-
tors, angiotensin receptor blockers, beta
blockers, statins).7 Although it is neces-
sary to remember the intrinsic limitations
of observational studies and the fact that
the theoretical benefits of such drugs must
always be weighed against their frequent
adverse effects, particularly in elderly
patients, these numbers suggest that there
are still ample margins for improving
the prevention of cardiovascular disease
among CKD patients.
The findings of this and previous stud-
ies have very important implications not
only for patients and their families, but also
for health-care providers, because the slow
rate of renal-function decline in the elderly
CKD population will lead in the near future
to a smaller number of patients requiring
long-term dialysis than expected, and pub-
lic-health programs can be adjusted on the
grounds that the need for dialysis centers
may be much less than foreseen only a
few years ago. Finally, without neglecting
the importance of therapeutic strategies
aimed at slowing down the progression of
renal damage in CKD patients, this study
provides further evidence that the nephro-
logical management particularly of elderly
CKD patients should first of all concentrate
on therapeutic options aimed at prevent-
ing or decreasing the development of the
cardiovascular abnormalities that cause the
death of these patients before they reach
the need for chronic dialysis.
2119
co m m e nta r y

REFERENCES Institute of Diabetes and Digestive and Kidney and iron) but is expected to improve qual-
1. Xue JL, Ma JZ, Louis TA, Collins AJ. Forecast of the Diseases, Bethesda, Maryland, USA. <www.usrds. ity of life. Is the incremental improvement
number of patients with end-stage renal disease org/adr.htm (2005).
in the United States to the year 2010. J Am Soc 5. Keith DS, Nichols GA, Gullion CM et al. in quality of life worth the extra costs of
Nephrol 2001; 12: 2753–2758. Longitudinal follow-up and outcomes among a anemia therapy? Previous work by Tonelli
2. Jungers P, Chauveau P, Descamps-Latscha B et
al. Age and gender-related incidence of chronic
population with chronic kidney disease in a large
managed care organization. Arch Intern Med 2004;
et al. suggests that it is not.2 The incre-
renal failure in a French urban area: a prospective 164: 659–663. mental cost-effectiveness ratio (ICER)
epidemiologic study. Nephrol Dial Transplant 1996; 6. Foley RN, Murray AM, Li S et al. Chronic kidney is commonly used to answer such ques-
11: 1542–1546. disease and the risk for cardiovascular disease,
3. Coresh J, Astor BC, Greene T et al. Prevalence of renal replacement, and death in the United States
tions. The ICER is calculated by division
chronic kidney disease and decreased kidney Medicare population, 1998 to 1999. J Am Soc of the incremental cost by the incremental
— — —
function in the adult US population: Third National Nephrol 2005; 16: 489–495. effectiveness (ICER =∆C /∆E, where ∆C is
Health and Nutrition Examination Survey. Am J 7. Hemmelgarn BR, Zhang J, Manns BJ et al. —
Kidney Dis 2003; 41: 1–12. Progression of kidney dysfunction in the the difference in mean costs and ∆E is the
4. United States Renal Data System. 2005 Annual community-dwelling elderly. Kidney Int 2006; 69: difference in mean effectiveness). Tonelli
Data Report. National Institutes of Health, National 2155–2161. et al.2 suggested that raising hemoglobin
from 9.5–10.5 g per dl to 11.0–12.0 g per
dl cost US$55 295 per quality-adjusted
see original article on page 2219 life-year gained. Increasing hemoglobin
further to the 12.0–12.5 g per dl range
was associated with an ICER of $613 015
Interpreting cost-effectiveness in per quality-adjusted life-year. Tonelli et
al.2 concluded that raising hemoglobin
dialysis: can the most expensive to the 11.0–12.0 g per dl range was cost-
effective, but higher targets were not. Of
be more expensive? course, before we can accept or reject a
new intervention, we need to know the
PA McFarlane1 maximum amount that the payer would be
willing to spend for the additional benefit.3
In this issue, Tonelli et al. describe the cost-effectiveness of In our anemia example, what is the maxi-
arteriovenous fistulae (AVF) screening, and conclude that such a mum that we are willing to pay to increase
program represents “good value for money.” Here, I examine the quality-adjusted survival? In other fields of
robustness of this conclusion by considering traditional definitions of medicine, health economists often point
to the cost of dialysis as a possible maxi-
acceptable cost-effectiveness and more pragmatic definitions, which
mum that we would be willing to spend
include consideration of the maximum amount that society would be to improve survival — a strangely circular
willing to pay for hemodialysis. argument for health economists working
Kidney International (2006) 69, 2120–2121. doi:10.1038/sj.ki.5001556 in nephrology. Other work has suggested
that the maximum willingness-to-pay
value for a quality-adjusted life-year is
In this issue, Tonelli and colleagues1 from be in keeping with a more fundamental between $60 000 and $160 000.4,5 By these
Alberta, Canada, attempt to identify approach to health economics, which is to standards, Tonelli and colleagues’ conclu-
whether screening of arteriovenous fistu- use a fixed set of resources (money, staff, sions about anemia therapy seem sound. In
lae (AVF) represents what they describe consumables, and so on) in the manner their current study, Tonelli et al.1 suggest
as “good value for money.” The concept of that produces the best patient outcomes. that it costs between $8 000 and $10 000
‘bang for the buck’ underlies many health- Those working in dialysis programs to avoid one AVF failure, and they con-
economic analyses. Typically, we consider should be aware that in-center hemodi- clude that this “may represent good value
this concept in the following context: ‘This alysis is often held up as the benchmark for money.” Unlike in the anemia exam-
new intervention will improve patient maximum that society is willing to pay ple, where empiric work had suggested a
outcomes but will cost more as well. Is the for a chronic life-saving intervention. possible maximum willingness to pay to
extra benefit worth at least as much as the With that in mind, let’s consider whether improve quality of life, no previous work
extra cost?’ This common application of the work of Tonelli et al.1 produces “good has suggested how much we are willing to
the ‘bang for the buck’ approach may not value for money” for dialysis programs by spend to save a fistula. Should we believe
both economic approaches. Tonelli and colleagues’ conclusions about
1Department of Medicine, University of Toronto,
An example of a traditional approach AVF screening? One solution may be to
Toronto, Canada
to determining value for money is anemia perform additional research to quantify
Correspondence: PA McFarlane, Department of
Medicine, University of Toronto, 61 Queen Street management. Targeting a higher than our willingness to pay in this area. Alterna-
East, 9th Floor, Toronto, Ontario M5C 2T2, Canada. usual hemoglobin level may cost more tively, we can look to a broader economic
E-mail: phil.mcfarlane@utoronto.ca (because of greater use of erythropoietin view of dialysis to provide answers.

2120 Kidney International (2006) 69