NSAID Nephropathy and COX-2
Kellie A. Goldsborough, MD
Resident Grand Rounds Presentation
Wake Forest University Baptist Medical Center
Department of Internal Medicine
November 2, 1999
HPI- JS is a 69 year old black male with a past medical history significant for dilated
cardiomyopathy, alcohol abuse, and gout who presented to the ED on 7/18 with a two week history
of nausea, anorexia and confusion. Of note, he was seen in the ED on 6/30 and given Indomethacin
for a gout exacerbation.
PMHx- Dilated cardiomyopathy; echo 11/97= EF of 20-25%
History of alcohol abuse
MEDS- benazepril 20mg q day ALL- NKDA
lanoxin 0.125mg q day
furosemide 40mg q day
K-Dur 20meq q day SH- tobacco, current ETOH
Indomethacin TID x 3 weeks
PE- T-98.2 P-79 BP-110/79 R-16 97% on RA
GEN- thin, bm, slow to respond, NAD
HEENT- PERRLA, EOMI
NECK- no bruits or JVD
HEART- RRR, w/o M/G/R
ABD- good bowel sounds, non-tender, non-distended
EXT- no clubbing, cyanosis, or edema
NEURO- overall depressed but no focal deficits
His laboratory exam was significant for a BUN/CRT of 147/4.7 (had been 13/1.1 in 6/99).
He was admitted to the renal service, the NSAID and ACE were discontinued. Work up included a
negative UA, renal ultrasound and a duplex negative for renal artery stenosis. His ACE was restarted
and he was eventually discharged from the hospital on 7/21 with BUN/CRT of 54/1.0. He was
advised to not take NSAIDS and was to follow up at RHC.
He did not keep his appointment, and eventually returned to the ED on 8/12 after taking a
large amount of an OTC "pain med", with a BUN/CRT of 230/18.7. His work up again revealed a
negative UA, and ultrasound. Additional lab tests included a negative SPEP/UPEP, ANCA and
antiGBM. With intermittent use of hemodialysis and a brief stay in the MICU, his renal function
eventually recovered. His ARF was presumed to be secondary to NSAIDS. He was eventually
discharged to a nursing home on 8/23 with a BUN/CRT of 4/0.7.
Why did this happen?
Could it have been predicted or avoided?
Would the substitution of a different NSAID or a COX-2 inhibitor produce any less risk for this kind of
Approximately 1- 5% of patients exposed to NSAIDS develop some type of nephrotic
syndrome warranting potential physician intervention. With the extensive use profile of NSAIDS as
analgesic, anti-inflammatory and anti pyretic agents, an enormous number of US citizens are at risk
for consequential kidney dysfunction. Approximately 1 in 7 patients with a rheumatological disorder
is likely to receive a prescription for an NSAID, and 1 in 5 US citizens report NSAID use for other
acute complaints. Thus, it is possible to estimate some type of renal abnormality in 500,000 to 2.5
million US citizens exposed to NSAIDS on a regular/intermittent basis per year. Use of NSAIDS may
spiral upward still with the aging of the population, increasing availability over the counter and with
the advent of new selective COX-2 inhibitors which will expand the number of patients who can
tolerate these medications. The renal toxicity of NSAIDS can be divided into several distinct clinical
1. Abnormalities in sodium, water and potassium homeostasis
2. Vasomotor acute renal failure
3. Nephrotic syndrome
4. Interstitial nephritis
5. Chronic renal failure most often due to papillary necrosis (acute or chronic)
Although these are the most common, there have been a multitude of renal lesions described,
these are listed in table 1.
Renal Lesions Found in Patients
Acute Tubular Necrosis
Acute Interstitial nephritis with or without nephrotic syndrome
Minimal change disease without interstitial nephritis
Chronic Renal Insufficiency and ESRD
Undefined or other lesions
Table 1. Renal Lesions found in patients taking NSAIDS
In this presentation, I plan to discuss the pathophysiology behind the forms of NSAID
nephropathy, focusing on the most common and predictable, the vasomotor acute renal failure
syndrome. I will outline risk factors for the development of NSAID nephropathy, and discuss the
differences within the traditional NSAID group as well as the COX-2 inhibitors with regard to
Types of Nephropathy-
D. Interstitial nephritis (IN)
There are at least 100 cases of biopsy proven interstitial nephritis known to be related to normal
doses of NSAIDS. This is sure to be an underestimation, because the phenomena, when looking at
NSAIDS only, is less predictable and less specific because it is also associated with a number of other
medications, such as beta lactams.
The mechanism for IN related to NSAIDS is unclear. It has been postulated to result from a
delayed hypersensitivity response to the NSAID. Other theories include inhibition of the
prostaglandin synthesis shunting arachidonic acid catabolism toward pro-inflammatory substances.
The patients afflicted are often elderly, predominately female, and taking NSAIDS for months. The
clinical manifestations include heavy proteinuria, increased RBC, and leukocytes in the microscopic
analysis, and a low fractional excretion of sodium. Treatment consists of withdrawing the NSAID,
and supportive care. The use of steroids has been debated. In most cases AIN and the nephrotic
syndrome resolve with the discontinuation of NSAIDS over a period ranging from weeks to a year.
However, in at least 2 studies, a high percentage of patients had permanent renal damage, compared
to the outcome of those suffering from other types of NSAID related lesions, the renal failure
reversibility estimated at 25%.
D. Nephrotic syndrome
Approximately 10% to 12% of patients developing renal lesions while receiving NSAIDS
have minimal change nephrotic syndrome. The typical picture is a female taking the drug for months
before proteinuria is discovered. Examinations of renal specimens by light microscopy show normal
glomeruli and electron microscopy shows fusion of the epithelial cell foot processes, common in
minimal change disease. Treatment, like for IN, consists of withdrawing the NSAID and supportive
care. Steroids are thought to be more beneficial in this population, based on the model of traditional
minimal change disease.
D. Renal Papillary Necrosis
This represents the least common and the most serious of the nephropathies, most often
resulting in end stage renal disease. It is usually seen in the setting of massive NSAID overdose in a
dehydrated patient. It presents either in an acute or chronic form.
a. Acute- the underlying pathophysiologic process appears to be ischemic
necrosis. Clinical circumstances of severe dehydration and massive NSAID
ingestion leads to elevated toxic NSAID metabolites within the papillae. The
presence of these metabolites also inhibits vasodilation. The result is papillary
b. Chronic- also known as 'analgesic nephropathy' and is present in approximately
2% of the hemodialysis population. It was initially reported in patients who
had repetitive daily ingestions of compound analgesic mixtures most often
containing aspirin and phenacetin. This was noted to produce a syndrome of
chronic renal failure induced by tubulointerstitial nephritis involving the
medulla and papillae. Removal of phenacetin from the market has not led to
the eradication of this syndrome and it has now been described in relation to
Pathophysiology of Vasomotor Acute Renal Failure
Vasomotor acute renal failure is the most common and predictable form of NSAID
nephropathy, and is directly related to prostaglandin synthesis suppression in the kidney.
Prostaglandins (PGE2, PGI1, PGF2) are unsaturated fatty acid compounds derived from 20 carbon
essential fatty acids like arachidonic acid. Prostaglandins function for the most part as local
hormones or autocoids to their respective organs. The formation of prostaglandins from arachidonic
acid is catalyzed by two highly homologous isoforms of the enzyme cyclooxygenase (COX-1 and COX-
2). The synthesis and release of the prostaglandins are influenced by numerous physiologic stimuli
including AII, vasopressin, catecholamines, and trauma. Interestingly, under baseline euvolemic
conditions, prostaglandin synthesis in the kidney is negligible and, as a result, these compounds play
little to no role in the minute to minute maintenance of renal function. Where these compounds
come to serve a major role is in the setting of a systemic or intrarenal circulation disturbance, such as
volume depletion. In this setting, renal blood flow is decreased. Responses to the hemodynamic
challenge include stimulation of the renin/angiotensin/aldosterone axis, which promotes
vasoconstriction and sodium and chloride reabsorption, and elevated sympathetic outflow which
further tends to increase vascular tone. Renin, angiotensin II and catecholamines also stimulate
prostaglandin synthesis as stated above. The role of the prostaglandins in this setting is to provide a
compensatory vasodilation of the renal afferent and efferent vasculature to ensure adequate blood
flow and preclude acute functional deterioration in the kidney. It is largely by blunting these
prostaglandin mediated counterregulatory mechanisms that NSAIDS compromise renal function.
Volume contraction, CHF, cirrhosis
adrenergic system renin-angiotensin which NSAIDS
norepinephrine angiotensin II response of renal
Renal prostaglandin synthesis
Obviously, since we have just discussed in detail the pathophysiology of NSAID induced
acute renal failure, there is an abundance of evidence describing the relationship. Based on the
mechanism noted above, one could postulate certain disease states which would place the patients at
risk for development of acute renal failure. Any physiologic state which results in a decreased
effective arterial blood volume would make renal perfusion more dependent on prostaglandins.
These disease states include true decrease in blood volume with dehydration, hemorrhage or
diuretic treatment, or perceived decrease in effective arterial volume such as with congestive heart
failure, cirrhosis, and nephrotic syndrome. In fact, the association between these disease states and
vasomotor acute renal failure has been reported in a number of observational and case series
articles. In addition, the risk was expanded to include anyone with a decreased GFR, such as people
with longstanding, stable chronic renal failure or the elderly, despite a normal hemodynamic status.
In chronic renal failure, the prostaglandins are thought to play an adaptive role in
minimizing the decline in global renal function by increasing the GFR in the surviving nephrons. In
patients with chronic renal insufficiency, an association has been found between increased urinary
excretion of PGE2 at baseline and risk for NSAID induced renal impairment. In other words, when the
compensatory level of PGE2 gets to a certain threshold, any inhibition at that point would
compromise renal function. However, other studies have found that deficient production of
prostacyclin (indicated by decreased 6 keto PGF1) at baseline is a truer marker for identifying
patients at risk. This is important because a popular endpoint in the studies coming up is urinary
prostaglandin trends. It is yet to be determined which urinary prostaglandin is of more importance
in predicting risk.
Another group that may be more vulnerable to the toxic effects of NSAIDS include the
elderly. Aging leads to several changes including: 1) a decrease in GFR 2) the renal vasculature
becoming less responsive to vasodilators 3) decreasing albumin levels which reduces protein
binding of the NSAID and result in a higher free drug concentration 4) decreased hepatic metabolism
Table 2 : Risk Factors for NSAID induced vasomotor renal failure
Decreased effective arterial volume Normal or increased effective arterial volume
Congestive Heart Failure Chronic Renal Failure
Nephrotic Syndrome Elderly age
Sepsis Contrast induced nephropathy
Hemorrhage Obstructive uropathy
Diuretic Therapy Cyclosporine use
Post op patients
Are All NSAIDS Created Equal?
Given that there are a number of patients at high risk for NSAID nephropathy, an NSAID with
less nephrotoxicity would be desirable. The most attention, up until the development of COX-2
inhibitors, had been toward sulindac in this regard.
In 1984, Bunning et al published a case series which consisted of three patients who developed acute
renal failure with indomethacin, ibuprofen and naproxen respectively with no other risk factors for
acute renal failure. They then were tried on sulindac during the same hospitalization and had no rise
3 Fig. 2
2 Increase in serum creatinine level
1.5 during NSAID therapy with rapid
1 return to baseline after therapy
0.5 d/cd, then no change in creatinine
0 with sulindac therapy
Their conclusions were that in contrast to other NSAIDS, sulindac has significantly less renal
toxicity. Their explanation for this had to do with the pharmacokinetics of sulindac. The drug taken
by mouth is actually a prodrug. It is converted to the active drug, the sulfide form, in the liver. The
sulfide is then irreversibly converted to a sulfone that is again inactive as a prostaglandin inhibitor.
Thus, it was determined that the kidney wasn't exposed to an actual prostaglandin inhibiting action,
therefore sparing renal function.
A. Mistry et al
Mistry wanted to look at sulindac under more controlled conditions, so they studied 9
patients, ages 35-45 with chronic renal insufficiency (clearance 24.7-54.6ml/min), and hypertension.
Exclusion criteria included CHF, GI disorders, bleeding disorders and allergies to NSAIDS. Men were
excluded from urinary prostaglandin measurements because of the high prostaglandin content of
prostatic secretions complicating the assay.
Fig. 3 protocol of Mistry et al, Clin. Science (1986) 70, 501-505
9 pts Baseline Sulindac Test values 24 hour urine collection
crt clear 25- values 200mg BID obtained for creatinine, daily
55 ml/min obtained X 9 days elytes, daily clearance,
HTN GFR, renal plasma blood
flow, urinary PG levels
in females only
Exclusion: CHF, bleed, GI
disorder, allergy to NSAID
During the administration of sulindac, there were statistically significant falls in
creatinine clearance and rise in creatinine which were reversible.
The change in excretion of prostaglandins (PGE2 and 6 ketoPGF1) was not significant.,
but was decreased
Table 3 Change in parameters with sulindac tx in patients with CRI and hypertension
baseline sulindac p
creatinine clearance 37 2.2ml/min/m2 342.2ml/min/m2 <0.02
creatinine 195.819.7mol/L 20818.5mol/L <0.02
urinary PGE2 47.213.6 ng/h 3515.7ng/h NS
Urinary 6keto PGF1 19.68.7 ng/h 13.3 ng/h NS
The significant renal impairment coupled with the hypertension made these patients
more susceptible to cyclooxygenase inhibition by sulindac.
Although no NSAID should be regarded as safe in this population, sulindac does seem to
affect renal function only marginally, and used cautiously, it may have advantages over
other NSAIDS in such patients.
Small study, only 9 days of treatment, no placebo group
Urinary prostaglandins only collected in 5 patients
The changes termed significant were small differences with large confidence intervals of
unknown practical significance (creatinine translation 2.2-2.4 mg/dL)
B. Whelton et al- sulindac
This was a prospectively randomized triple crossover study to compare sulindac to other
traditional NSAIDS. The study population included 12 women with chronic renal failure (creatinine
130-270mol/L). The women were randomized to Ibuprofen 800mg. TID, Piroxicam 20 mg. q day or
Sulindac 200 mg. BID for 11 days. The endpoints followed included serum creatinine, effective renal
plasma flow, GFR, urinary prostaglandin excretion, and drug concentrations.
Ibup 800x11d End points followed:
Crt 1.5- Triple serum creatinine, effective renal plasma flow,
Pirox 20 x 11d xover GFR, urinary prostaglandins, drug concentrations
Fig. 4 Protocol of Whelton et al, Ann Int. Med. 1990
Ibuprofen was withdrawn on day 8 because of increased creatinine in two
patients and hyperkalemia in one.
All patients tolerated treatment with sulindac and piroxicam, but there was
a statistically significant increase in creatinine with sulindac (p<0.05)
Statistically significant suppression of urinary prostaglandin activity was
observed for all three groups.
day 2 day 5 day 12
Fig 5. Percent of baseline urinary PGE2 with treatment of different NSAIDS
This cautioned the extrapolation of previous reports that sulindac was renal sparing.
This population of patients had more severe renal insufficiency and stayed on therapy
for longer in contrast to the studies which showed renal sparing.
Therefore, they determined that the extent of sulindac sulfide accumulation may be a
determining factor in the development of renal impairment.
As a guideline, they identified a creatinine of 180 mol/L or 2 mg/dL as the cutoff for a
risk factor. They also recommend checking this type of patient's renal function 7-10
days after starting any NSAID, including sulindac.
Small study of only women, no placebo group
The patients who developed renal insufficiency are not reflected in the final data
C. Eriksson et al
This was a double blind, crossover design comparing sulindac to naproxyn. The 9 patients
studied had a rheumatologic disease and chronic renal insufficiency (mean clearance 53
ml/min/m2). Some had cardiovascular disease but were otherwise healthy. This experiment was
highly supervised with patients hospitalized. The parameters followed were: electrolytes, urine
volume, creatinine, GFR, RPF, urinary and serum aldosterone and urinary 6 keto PGF 1.
9 pts Sulindac x 7d Sulindac x7d Electrolytes, urine
renal insuff volume, creatinine,
placebo 7d Placebo 7d
clearance GFR, RPF, urinary 6
53ml/min keto PGF1
rheum disease Naproxen x7d Naproxen 7d
Fig. 6 Protocol for Eriksson et al, AM J of Med, vol 89, Sept, 1990
RESULTS / CONCLUSIONS:
Treatment with naproxen significantly decreased the GFR, RPF and the excretion of 6-
ket-PGF1, while sulindac had no such effect.
There was no change in creatinine with either treatment
Short-term treatment with sulindac does not appear to suppress the GFR and RPF in
patients with reduced renal function in contrast to other NSAIDS.
Table 4 Changes in variables during treatment with sulindac vs. naproxen
control naproxen p naproxen control sulindac
creatinine 94 7mol/L 98 8 mol/L NS 97 7mol/L 108 10mol/L
GFR 35 7ml/min 26 4ml/min <0.05 314 ml/min 295ml/min
RPF 23549 ml/min 18727ml/min <0.05 21123ml/min 19729ml/min
urine 6 keto PGF1 4.40.9ng/hr 1.30.2ng/hr <0.01 4.20.7ng/hr 5.60.9ng/hr
9 patients, treatment period only 7 days
7 day washout period may not have been adequate
no placebo group
Table 5 Summary of studies involving sulindac as a renal sparing NSAID
characteristics creatinine clearance GFR Urinary PGs
Mistry 9 pts, crtclearance 25-55,HTN NM No
9 days of therapy
Whelton 12 women, CRI with crt 1.5-3.0 in NM No in all
11 days of therapy ibuprofen difference groups
sulindac, ibuprofen, piroxicam and sulindac
Eriksson 9 pts, crt clearance 53 ml/min No in in in
7 days of therapy
difference naproxen naproxen naproxen
sulindac and naproxen
Since 1971, it has been generally accepted that the mechanism for both therapeutic anti-
inflammatory, analgesic, and antipyretic actions and the common deleterious effects of ASA like
drugs is mediated through their inhibition of COX, the rate limiting enzyme in the synthesis of
prostaglandins. We now know that COX exists in 2 isoforms known as COX-1 and COX-2. The 2
isoforms have different structures and functions. The so called "constitutive"
isoform COX-1 is involved in processes such as the production of thromboxane A 2, the production of
PGE2 in the kidneys and the production of prostacyclin which is both anti thrombogenic and, in the
gastric mucosa, cytoprotective. COX-2 was thought to be
the "inducible" form which is turned on by inflammatory stimuli such as cytokines and
lipopolysaccharides. Therefore, in theory, an NSAID which displays preferential COX-2 inhibition
would be expected to have potent anti-inflammation effects while sparing the patient from adverse
However, the classification of constitutive and inducible isoforms is misleading because COX-
1 has been found to be induced and COX-2 is constitutively expressed in some tissues including the
prostate, brain, lung, intestine and kidney. In a study by Komhoff et al in 1997, expression of COX-2
immunoreactive protein could be localized to the endothelial and smooth muscle cells of arteries and
veins in the human kidney suggesting that COX-2 might contribute to the regulation of glomerular
hemodynamics. But, does the presence of the COX-2 enzyme guarantee a vital role? Therefore, there
was a need to perform studies to look at renal function and COX-2 inhibitor administration.
Table 5 Influence on COX-1 and COX-2 activity of guinea-pig peritoneal macrophages for different NSAIDS
NSAID COX-1 IC50(mmol/L) COX-2 IC50 (mmol/L0 Ratio COX-2/COX-1
Flurbiprofen 15 4760 317
Indomethacin 0.21 6.4 30
Piroxicam 5.3 175 33
Meloxicam 5.8 1.9 0.33
SC-236 17.8 0.01 0.00056
A. Stichtenoth et al
The first marginally selective COX-2 inhibitor available for study was meloxicam. It has
never been available in the US. It displays less specific COX-2 selectivity than the newer drugs, but
more than any of the traditional NSAIDS had at the time. Stichtenoth, in 1997, looked at the
administration of meloxicam to healthy women and its effects on renal function. It was a
randomized, cross over design comparing indomethacin and meloxicam.
Meloxicam 7.5 qday x 6d
age 18-45 Endpoint: 24 hour
5 day wash urinary excretion
out of PGE2 and PGE-
Exclusion: acute or
chronic illness, ulcers, Indomethacin 25 TID x 3d M
pregnancy, lactation, any
meds xcept OCPs Fig. 7 Protocol of Stichtenoth, et al, J of Invest Med., 1997
RESULTS / CONCLUSIONS:
There was a reduction of urinary PGE2 by 43 % (p<0.05) and a reduction of PGE-M by
36% (p<0.001) with indomethacin when compared to baseline controls, indicating a
true inhibition of prostaglandin synthesis.
With meloxicam, the urinary PGE2 was decreased by 13 % which did not show statistical
significance, and the PGE-M was reduced by only 22% (p<0.05).
There was statistical significance reached between indocin and meloxicam with regard
to PGE2 excretion.
Urinary PGE2 excretion, reflecting physiologic renal PGE2 production in healthy
volunteers, was not inhibited by meloxicam, while indomethacin caused pronounced
80 Fig. 8 urinary PGE2
levels in healthy women
40 and meloxicam
Again, only women in study
No placebo group
Measured surrogate endpoints only
B. Bevis et al
Bevis et al performed an open label multicenter study with meloxicam. He was interested in the
population of patients with baseline renal impairment. He gathered 25 patients who had arthritic
disease and mild renal impairment (creatinine clearance 25-60 ml/min) and observed them while
administering meloxicam 15 mg q day for 28 days after a washout period.
25 patients Fig. 9 protocol of Bevis et al, Brit J of Rheum, 1996
renal impairment with washout 4 days Meloxicam treatment
creatinine clearance of 25- 15 mg. qday x 28d
60 ml/min, arthritis
Exclusion: severe cardiac, pulm, GI,
Hepatic or neuro dz, asthma,
ulcers, hemodialysis, tx with Endpoints at day 14, 21, 28, 35
anticoag, lithium or ACE, abnml Creatinine clearance, creatinine,
baseline labs electrolytes
RESULTS / CONCLUSIONS:
There was no significant difference in the mean creatinine clearance from baseline
compared to 14, 21, and 28 days.
There were also no rises in BUN or potassium levels throughout the testing period.
They determined that the preferential selectivity for one COX isomer over another is
implicated in the differing incidence of renal toxicity observed between various NSAIDS,
and determined the COX-2 selective inhibitor to have minimal renal adverse effects.
Didn’t measure the urinary prostaglandins which appears to be a surrogate marker for
No placebo group, or comparison
D. Bosch-Marce et al
Finally, the more powerful COX-2 inhibitors are available. What about the effects in patients
who have normal renal function but have a disease state which predisposes them to NSAID
nephropathy? No studies on humans, however, in 1999, Bosch-Marce performed a study on rats with
cirrhosis and ascites and administered NSAIDS including a COX-2 selective inhibitor to look at effects
on renal function.
supratherapeutic Parameters followed:
ketorolac urine volume, GFR, renal plasma
22 rats induced
flow, urinary PGE2 and urinary 6
cirrhosis and ascites
SC-236 Fig. 10 protocol of Bosch-Marce et al, Gastroenterology, 99
Table 5 changes in parameters in rats with ascites
60 minutes after administration of ketorolac or SC-236
base line ketorolac baseline SC-236
urine vol 26.55.1 10.12.4* 27.54.1 19.34.0
GFRml/min 3.10.5 1.60.4* 3.40.4 2.70.4
RPFml/min 18.02.5 12.12.7 20.43.2 16.93.7
* = p<0.01, = p<0.05
60 min after 60 min after
400 dose 200 dose
ketorolac SC-236 ketorolac SC-236
Fig. 11 changes in 6 keto PGF1 (left) and PGE2 (right) levels 60 minutes after NSAID administration
RESULTS / CONCLUSIONS:
While ketorolac administration resulted in a significant decrease in urine volume, GFR,
renal plasma flow and urinary prostaglandin excretion, SC-236 failed to reach statistical
significance in any of these categories.
As part of the study, an assay looking for COX-1 and COX-2 activity was performed, and
constitutive expression was found in the rat kidney
They concluded that although rats with cirrhosis and ascites showed constitutive COX-2
mRNA expression, selective inhibition of this isoform was apparently devoid of any
significant effect on renal function.
To this point, these are all of the trials in publication in order to determine the true renal
effects of COX-2 inhibitors. The following are abstracts from additional trials obtained from Pfizer
and not in publication yet, the following two abstracts are going to be presented at the ASN meeting
later this week.
D. Whelton et al - elderly
This is a study funded by the drug manufacturer of celecoxib. It is a single center, single-
blind, randomized, crossover study comparing the effects of celecoxib and naproxen on renal
function and urinary PGE2 and 6keto PGF1 excretion in 29 healthy elderly subjects with a GFR of
Celecoxib 200BID x 5d Fig 12 protocol of
Then 400 BID x 5d Whelton et al-
29 healthy elderly celecoxib and
GFR 67-127 ml/min 7 day elderly patients to
washout look at renal
Naproxen 500 BID for function
There was a significantly greater reduction in GFR with naproxen (-7.53ml/min)as
compared to celecoxib (-1.11ml/min) (p=0.004)
Urinary PGE2 was significantly reduced from baseline with both celecoxib and naproxen
by 65% and 76% respectively (p<0.05 and p<0.042)
Urinary keto PGF1 was reduced to undetectable levels following the administration of
both celecoxib and naproxen.
Celecoxib does reduce urinary prostaglandins but does not affect GFR in healthy elderly
The patients in this study did not have very significant renal dysfunction, as compared to
the usual elderly population; can it be applied to patients with renal dysfunction?
No placebo group, 7 days washout sufficient?
Funded by drug company
E. Whelton et al- renal insufficiency
The makers of celecoxib wanted to test their drug in a population of patients at risk. This
study was necessary because the patients in the previous study did not have significant renal
dysfunction. It is a multicenter, double blind, randomized; placebo controlled parallel group study
comparing the renal effects of celecoxib and naproxen in 75 patients with stable chronic renal
Celecoxib 200 BID Fig 13 protocol of
x 7 days Whelton et al
study of celecoxib
75 patients in renal
Naproxen 500 BID PGE2, 6 keto PGF,
GFR 40-60ml/min insufficiency
x 7 days GFR
crt 1.3-3.0 mg/dL patients
Placebo x 7 days
There were no significant differences noted in GFR between the groups.
Both active agents reduced urinary prostaglandin output, but the effect of naproxen
consistently exceeded that of celecoxib with 88% and 82% reduction with naproxen and
47% and 48% reduction with celecoxib. However, the differences between the two
treatments failed to reach statistical significance
Although these studies demonstrate a reduction in urinary prostaglandins, they do not show
a decrease in the GFR or an increase in creatinine related to these medicines. However, based on
previous studies as well as the mechanism, this seems to be the next logical progression. Are there
any studies linking COX-2 inhibitors to renal dysfunction?
Local Case Reports of COX 2 inhibitor related nephropathies- Deterding
AM is a 65 yo wf with past medical history significant for CML (dx 3/98), DM, HTN and
previous episode of acute renal failure thought to be due to tumor lysis syndrome with resolution
and baseline BUN/crt of 23/1.2. She had been started on Vioxx on 8/13 and she presented to the
ER on 8/20 with a BUN/crt of 106/8.8. She was also on accupril and lasix but had been for years
without incident. While hospitalized, her work up included a negative urinalysis, ultrasound and
duplex. The Vioxx and ACE were discontinued and her renal failure resolved. She was discharged 4
days later with a BUN/crt of 26/0.9.
GB is a 72 yo wf with past medical history significant for obesity, DM, and chronic venous
insufficiency who had a prolonged hospitalization (5/28-6/17) for ?PE, pneumonia and bacteremia,
all of which improved with treatment. She was started on lasix on 6/7, an ace inhibitor on 6/9, then
on celecoxib on 6/12. Labs on 6/16 showed normal renal function She was then transferred to the
TCU. The next lab check was on 6/23, which found BUN/crt of 25/2.9, then 29/3.6 on 6/25, and
29/3.9 on 6/26. The celecoxib and ace were discontinued, work up included negative renal
ultrasound and duplex, and a urinalysis showing 1-5 WBC and few bacteria. Renal function improved
after stopping the celecoxib, on 10/7 it was 17/1.0.
These two cases are very suggestive of COX-2 related nephropathy, with the easy
reversibility and the negative work ups. However, there were other variables involved, most notably
the diuretics and ace inhibitors. With no evidence of renal artery stenosis, it is unlikely that the ace
inhibitors were responsible for the acute renal failure, but this cannot be verified without other
1. NSAID nephropathy is a significant health issue and the prevalence may increase in the future
due to aging of the population, increased availability, and the new COX-2 inhibitors.
2. Vasomotor acute renal failure is the most common and predictable NSAID nephropathy and also
the most easily avoided by careful administration of NSAIDS.
3. Risk factors for acute renal failure related to NSAIDS are: volume depletion, congestive heart
failure, cirrhosis, nephrotic syndrome, chronic renal insufficiency and advanced age.
4. When looking at traditional NSAIDS, sulindac may pose less of a risk for acute renal failure if
used for a short course.
5. The COX-2 enzyme is constitutively expressed in the kidney, and COX-2 inhibitors, initially
promising, appear to cause inhibition of prostaglandin synthesis and limited demonstration of
acute renal dysfunction. However, this may be to a lesser degree than traditional NSAIDS, more
studies are needed to delineate.
6. No NSAID can be regarded as completely renal sparing; we must use these medicines more
judiciously with careful monitoring of renal function in those who are at risk.
1. Vane, J.R., "Introduction: Mechanism of Action of NSAIDS", British Journal of Rheumatology, 35
(suppl. 1):1-3, 1996.
2. Whelton, Andrew, "Nephrotoxicity of Nonsteroidal Anti-inflammatory Drugs: Physiologic
Foundations and Clinical Implications", The American Journal of Medicine, Volume 106 (5B),
May 31, 1999.
3. Kleinknecht, Dieter, "Interstitial Nephritis, the Nephrotic Syndrome, and Chronic Renal Failure
Secondary to Nonsteroidal Anti-Inflammatory Drugs", Seminars in Nephrology, Vol. 15, No. 3, pp.
228-235, May, 1995.
4. Palmer, B., "Clinical Acute Renal Failure With Nonsteroidal Anti-Inflammatory Drugs", Seminars
in Nephrology, Vol. 15, No. 3, pp. 214-227, May, 1995.
5. Clive, D. M., "Renal Syndromes Associated with Nonsteroidal Antiinflammatory Drugs", New
England Journal of Medicine, Vol. 310, No. 9, pp. 563-572, March, 1984.
6. Sandler, D., "Nonsteroidal Anti-Inflammatory Drugs and the Risk for Chronic Renal Disease",
Annals of Internal Medicine, Vol. 115, No. 3, pp. 165-172, August, 1991.
7. Bunning, R. "Sulindac: A Potentially Renal-Sparing Nonsteroidal Anti-inflammatory Drug", JAMA,
Vol. 248, No. 21, pp. 2864-2867, December, 1982
8. Mistry, C., "Effects of sulindac on Renal Function and Prostaglandin synthesis in patients with
moderate chronic renal insufficiency", Clinical Science, Vol. 70, pp. 501-505, 1986.
9. Whelton, A., "Renal Effects of Ibuprofen, Piroxicam, and Sulindac in Patients with Asymptomatic
Renal Failure", Annals of Internal Medicine, Vol. 112, pp. 568-576, 1990.
10. Eriksson, L., "Effects of Sulindac and Naproxen on Prostaglandin Excretion in Patients with
Impaired Renal Function and Rheumatoid Arthritis", The American Journal of Medicine, Vol. 89,
pp. 313-321, September, 1990.
11. Engelhardt, G., "Pharmacology of Meloxicam, A New Nonsteroidal Antiinflammatory Drug with
an Improved Safety Profile Through Inhibition of COX-2", British Journal of Rheumatology, Vol.
35(suppl. 1):4-12, 1996.
12. Distel, M., "Safety of Meloxicam: A Global Analysis of Clinical Trials", British Journal of
Rheumatology, Vol 35(suppl 1):68-77, 1996.
13. Komhoff, M., "Localization of Cyclooxygenase-1 and 2 in Adult and Fetal Human Kidney:
Implication for Renal Function", American Journal of Physiology, Vol 272, F460-F468, 1997.
14. Stichtenoth, D., "Effects of Meloxicam and Indomethacin on Cyclooxygenase Pathways in Healthy
Volunteers", Journal of Investigative Medicine, Vol. 45, NO. 2, pp. 44-49, February, 1997.
15. Bevis, P., "An Open Study to Assess the Safety and Tolerability of Meloxicam In Subjects with
Rheumatic Disease and Mild Renal Impairment", British Journal of Rheumatology, Vol. 35(suppl.
16. Bosch-Marce, M., "Selective Inhibition of Cyclooxygenase 2 Spares Renal Function and
Prostaglandin Synthesis in Cirrhotic Rats with Ascites", Gastroenterology, Vol. 116:1167-1175,