Professional Documents
Culture Documents
Carlos Cantú-Brito a
a
Department of Neurology and Psychiatry, Instituto Nacional de Ciencias Médicas y Nutrición ‘Salvador Zubirán’,
Mexico City, b Department of Neurology, Hospital Civil de Guadalajara ‘Fray Antonio Alcalde’, Guadalajara,
c
Department of Neurology, Universidad Autónoma de Guadalajara, Zapopan, d Stroke Clinic, Instituto Nacional
de Neurología y Neurocirugía, Mexico City, e Department of Neurology, Hospital General de Culiacán, Culiacán,
f
Department of Neurology, Hospital Regional ‘Dr. Valentín Gómez Farías’, ISSSTE, Zapopan, g Department of Neurology,
Hospital Ángeles de Querétaro, Querétaro, h Medical Research Area, Sanofi-Aventis, Mexico City, Mexico
Key Words (362.8 ± 220.0 μmol/l). Compared with cases with higher
Chronic renal disease · Creatinine · Small-vessel disease, SUA levels at hospital admission, patients with ≤4.5 mg/dl
Mexico · Stroke · Urate · Uric acid (≤267.7 μmol/l; the lowest tertile of the sample) had more
cases of a very good 30-day outcome (30.5 vs. 18.9%, re-
spectively; p = 0.004). SUA was not associated with mortal-
Abstract ity or functional dependence (mRS >2) at 30 days, or with
Background: Current evidence shows that uric acid is a po- any outcome measure at 3, 6 or 12 months poststroke. After
tent antioxidant whose serum concentration increases rap- adjustment for age, gender, stroke type and severity (NIHSS
idly after acute ischemic stroke (AIS). Nevertheless, the re- <9), time since event onset, serum creatinine, hypertension,
lationship between serum uric acid (SUA) levels and AIS diabetes and smoking, a SUA ≤4.5 mg/dl (≤267.7 μmol/l)
outcome remains debatable. We aimed to describe the was positively associated with a very good short-term out-
prognostic significance of SUA in AIS. Methods: We studied come (odds ratio: 1.76, 95% confidence interval: 1.05–2.95;
463 patients (52% men, mean age 68 years, 13% with glo- negative predictive value: 81.1%), but not at 3, 6 or 12
merular filtration rate <60 ml/min at hospital arrival) with months of follow-up. When NIHSS was entered in the mul-
AIS pertaining to the multicenter registry PREMIER, who had tivariate model as a continuous variable, the independent
SUA measurements at hospital presentation. Multivariate association of SUA with outcome was lost. Compared with
models were constructed to analyze the association of SUA cases with higher levels, patients with SUA ≤4.5 mg/dl
with functional outcome as assessed by the modified Rankin (≤267.7 μmol/l) were more frequently younger than 55
scale (mRS) at 30-day, 3-, 6- and 12-month follow-up. A mRS years, women, with mild strokes, with normal serum creati-
0–1 was regarded as a very good outcome. Results: Mean nine and fewer had hypertension. The time since event on-
SUA concentration at hospital arrival was 6.1 ± 3.7 mg/dl set to hospital arrival was not significantly associated with
Serum Uric Acid and Stroke Outcome Cerebrovasc Dis 2013;35:168–174 169
DOI: 10.1159/000346603
Table 1. Main characteristics of the patients according to 30-day outcome
gender, renal function, time since stroke onset, stroke severity, mg/dl (362.8 ± 220.0 μmol/l), higher in men than in wom-
type of stroke) were taken into account. Multivariate analyses were en [6.6 ± 3.9 vs. 5.5 ± 3.5 mg/dl (392.6 ± 232.0 vs. 327.1 ±
constructed by forward stepwise binary logistic regression. Input
variables were those that resulted significantly associated with AIS 208.2 μmol/l); p = 0.002] (fig. 1).
outcome in bivariate analyses. Adjusted odds ratios (OR) with 95% At 30 days after AIS, a very good outcome occurred
confidence intervals (CI) are provided. The fitness of the models more frequently in patients with SUA ≤4.5 mg/dl (≤267.7
was evaluated by using the Hosmer-Lemeshow goodness-of-fit μmol/l; the lowest tertile of the sample) than in those
test, which was considered as reliable if p > 0.2. All p values are with higher levels (30.5 vs. 18.9%, respectively; p = 0.004;
two-sided and considered significant when p < 0.05. SPSS v 17.0
software was used for all statistical calculations. univariate Mantel-Haenszel OR: 1.88, 95% CI: 1.21–2.92).
Neither low nor high SUA levels were associated with
stroke mortality or functional dependence (mRS >2, or
mRS >3) at 30-day follow-up. Moreover, we could not
Results find any significant association between SUA and any
outcome at 3-, 6-, or 12-month follow-up, either in uni-
A total of 463 patients were analyzed (52% men, mean variate or in multivariate analyses.
age 68 years; range 21–104 years) (table 1); 61 (13%) pa- In a multivariate analysis adjusted for relevant cofac-
tients had glomerular filtration rate <60 ml/min at hospi- tors (table 2), SUA ≤4.5 mg/dl (≤267.7 μmol/l) was asso-
tal arrival. Mean SUA at hospital admittance was 6.1 ± 3.7 ciated with 30-day very good outcome (mRS = 0–1). How-
60
Furthermore, variables significantly associated with low
40 SUA levels were: female gender, young age, normal serum
creatinine and (inversely) hypertension (table 3).
20
0 Discussion
SUA first tertile 68$VHFRQGWHUWLOH 68$WKLUGWHUWLOH
(PJGO (²PJGO (!PJGO
In this study we found that a low SUA at hospital ad-
mission is modestly associated with a very good short-
Fig. 1. SUA tertiles as a function of gender.
term outcome. However, we could not demonstrate the
opposite, that severe strokes or adverse outcomes are as-
sociated with higher concentrations of SUA. Our results
Table 2. Factors positively associated with a very good outcome may contrast with some previous findings [9, 17, 20, 21,
(modified Rankin scale 0–1) at 30-day follow-up: a multivariate
29, 30], but are in line with the report of Nardi and Milia
logistic regression model*
[31] showing that a low SUA is associated with an excel-
Variable Multivariate odds p value lent functional state after AIS. Moreover, the factors that
ratios (95% CI) we found as associated with SUA concentrations have
NIHSS <9 points 6.99 (4.14–11.79) <0.001 also been described in similar settings [31]. Nonetheless,
Age <55 years 2.10 (1.18–3.74) <0.012 although from a different perspective, our findings sup-
Cardioembolic stroke 0.41 (0.19–0.88) <0.023 port the concept that SUA elevation is proportional to
SUA ≤4.5 mg/dl (≤268 μmol/l) 1.76 (1.05–2.95) <0.031 the magnitude of the brain ischemia [12, 13, 18], with low
* Hosmer-Lemeshow test for goodness of fit in final step of the SUA levels indicating a good outcome possibly through
regression model: χ2 = 1.85, 6 d.f., p = 0.933. Only variables sig- mild strokes. The present results offer relevant informa-
nificantly associated with mortality are shown. Model adjusted for tion on the factors that may confound and partially ex-
gender, chronic renal failure, diabetes, hypertension, dyslipidemia, plain the variation of SUA among AIS cohorts.
atrial fibrillation, lacunar stroke, undetermined stroke, admission The time of blood sampling appears to be a crucial issue
blood glucose <150 mg/dl (<8.3 mmol/l), admission serum creati-
nine <1.2 mg/dl (<106 μmol/l), hospital arrival <24 h since stroke that may explain much of the variation of results among
onset, previous cerebral infarction and body mass index >30. Only publications [9, 12, 13]. If SUA is a consumptive and rap-
variables significantly associated with a score of 0 to 1 in the mod- idly changing marker of the antioxidant response elicit by
ified Rankin scale are shown in this table. SUA association with the brain ischemia, then higher SUA levels should be ob-
short-term outcome showed significance only when dichotomised served during certain moments in larger strokes to offer a
NIHSS was considered instead of a continuous variable.
metabolic response for scavenging oxygen free radicals ex-
cessively produced during blood deprivation. Patients fail-
ing to mount such a reaction may have a bad outcome. In
ever, the significance of this association was lost when NI- the end, large strokes may originate in part the SUA re-
HSS was entered as a continuous variable. SUA ≤4.5 mg/ sponse, but may also be the consequence of a failing anti-
dl (≤267.7 μmol/l) had a sensitivity, specificity, PPV, oxidant reaction. This subject needs more exploration in
NPV, LR+ and LR– for predicting a very good 30-day out- basic and clinical studies. The considerable debate on the
come, of 47.7% (95% CI: 38.4–57.0), 67.4% (95% CI: 62.4– significance of SUA as a marker of AIS outcome may be due
72.1), 30.5% (95% CI: 24.1–37.9), 81.1% (95% CI: 76.2– to different interpretations of the same phenomenon [19].
85.1), 1.46 (95% CI: 1.14–1.87) and 0.78 (95% CI: 0.64– Cells and tissue preparations exposed to hypoxia/isch-
0.94), respectively. The time since AIS onset to hospital emia exhibit an increased expression of xanthine oxidase,
arrival was not significantly associated with AIS severity the rate-limiting enzyme in the conversion of hypoxan-
(scoring of the NIHSS), or SUA levels (fig. 2); neverthe- thine to xanthine and xanthine to uric acid [32]. Xanthine
Serum Uric Acid and Stroke Outcome Cerebrovasc Dis 2013;35:168–174 171
DOI: 10.1159/000346603
p for trend = 0.74 p for trend = 0.19
100 100
80 80
32.0
36.8 34.9
Frequency (%)
Frequency (%)
39.7
60 65.1 68.0 60 43.4
63.2 49.7
60.3
56.6
40 40 50.3
20 20
0 0
NIHSS NIHSS NIHSS SUA first SUA second SUA third
<9 points 9–18 points >18 points tertile tertile tertile
a Hospital arrival in <24 h Hospital arrival in ≥24 h b (≤4.5 mg/dl) (4.6–6 mg/dl) (>6 mg/dl)
Fig. 2. Hospital arrival according to stroke severity by the NIHSS (a), and SUA tertiles (b).
29.3
Frequency (%)
36.5 32.5
60 30 26.1 29.4 30.1
32.5
32.5
40 62.2 63.4
33.3 34.9
66.8 20
37.8 36.6
20 33.2 10 SUA >6 mg/dl
Fig. 3. Stroke severity by the NIHSS as a function of SUA tertiles, dividing the NIHSS distribution with wider (a) and closer (b) score
ranges.
oxidase is the only enzyme capable of catalyzing the for- Here we could not demonstrate a particular pattern of
mation of uric acid in humans. In other mammals, urate SUA concentration as a function of time from stroke onset
oxidase can metabolize uric acid to allantoin, a potent an- to hospital arrival, possibly because patients with mild
tioxidant, but this enzyme activity is lost in primates. It is strokes tended to arrive later than patients with profound
possible that uric acid production may have evolved as a neurologic deficits. However, although SUA levels were
compensatory mechanism in primates that cannot pro- measured at hospital admittance, we did not standardize
duce other potent organic antioxidants [33]. As such, uric precisely the moment of blood sampling or laboratory
acid production by means of xanthine oxidase activity is analyses, and no serial SUA measurements were under-
possibly the most potent acute antioxidant mechanism in taken; therefore, our data set is not adequate to evaluate the
response to ischemia in humans, and hence, it represents SUA time response as a function of stroke severity. This is
a marker of tissue infarction [22, 32, 33]. However, a lim- a very interesting topic to be examined in future studies.
itation of the models that examine the SUA systemic reac- The main limitation of this study is the sample size that
tion to focal ischemia (i.e. AIS) is that uric acid produc- might not be large enough to detect a more robust differ-
tion in situ may not be adequately measured with periph- ence between SUA levels across the whole range of stroke
eral blood sampling. severity, and to find that patients with severe strokes and
with higher SUA levels arrive earlier than their counter- AIS outcome. In other words, SUA is an acute indicator
parts. No volumetric analyses were performed to demon- of stroke severity if this response is proportional to the
strate that SUA is associated with the size of the cerebral infarction size. This idea should be consistently con-
necrosis, and no serial measurements were performed to firmed in further studies.
demonstrate that SUA levels change over time, as a func-
tion of stroke volume, clinical severity, renal function, age
and gender. Also, the previous use of thiazides and other Acknowledgements
uric acid-modifying drugs was not registered in our
The PREMIER study received unrestricted financial support
study, although indeed, none of the patients was treated
from conception to execution by Sanofi, Mexico. The pharmaceu-
with thiazides during the poststroke hospitalization. tical company was involved in the design of the study, but had no
In conclusion, these findings support the hypothesis role in the selection of patients, data analysis, the preparation of
that SUA is more a marker of the magnitude of the cere- this article or the decision for submission to publication.
bral infarction than a strong independent predictor of
References
1 Lehto S, Niskanen L, Rönnemaa T, Laakso M: 3 Holme I, Aastveit AH, Hammar N, Jungner I, 5 Kawai T, Ohishi M, Takeya Y, Onishi M, Ito
Serum uric acid is a strong predictor of stroke Walldius G: Uric acid and risk of myocardial N, Yamamoto K, et al: Serum uric acid is an
in patients with non-insulin-dependent dia- infarction, stroke and congestive heart failure independent risk factor for cardiovascular
betes mellitus. Stroke 1998;29:635–639. in 417,734 men and women in the Apolipo- disease and mortality in hypertensive pa-
2 Koton S, Howard SC, Warlow CP, Murphy protein MOrtality RISk study (AMORIS). J tients. Hypertens Res 2012;35:1087–1092.
MF, Rothwell PM: Serum urate predicts long- Intern Med 2009;266:558–570. 6 Waring WS: Uric acid: an important antioxi-
term risk of acute coronary events in women 4 Kim SY, Guevara JP, Kim KM, Choi HK, dant in acute ischaemic stroke. QJM 2002;95:
after a transient ischaemic attack and stroke. Heitjan DF, Albert DA: Hyperuricemia and 691–693.
Cerebrovasc Dis 2008;26:517–524. risk of stroke: a systematic review and meta-
analysis. Arthritis Rheum. 2009;61:885–892.
Serum Uric Acid and Stroke Outcome Cerebrovasc Dis 2013;35:168–174 173
DOI: 10.1159/000346603
7 Yang T, Chu CH, Bai CH, You SL, Chou YC, 17 Miedema I, Uyttenboogaart M, Koch M, Kre- 25 Logallo N, Naess H, Idicula TT, Brogger J,
Chou WY, et al: Uric acid level as a risk mark- mer B, de Keyser J, Luijckx GJ: Lack of asso- Waje-Andreassen U, Thomassen L: Serum uri
er for metabolic syndrome: a Chinese cohort ciation between serum uric acid levels and acid: neuroprotection in thrombolysis. The
study. Atherosclerosis 2012;220:525–531. outcome in acute ischemic stroke. J Neurol Bergen NORSTROKE study. BMC Neurol
8 Zhang Z, Bian L, Choi Y: Serum uric Acid: a Sci 2012;319:51–55. 2011;11:114.
marker of metabolic syndrome and subclini- 18 Seet RC, Kasiman K, Gruber J, Tang SY, 26 Cantú-Brito C, Ruiz-Sandoval JL, Murillo-
cal atherosclerosis in Korean men. Angiology Wong MC, Chang HM, et al: Is uric acid pro- Bonilla LM, Chiquete E, León-Jiménez C,
2012;63:420–428. tective or deleterious in acute ischemic stroke? Arauz A, et al: Acute care and one-year out-
9 Chamorro A, Obach V, Cervera A, Revilla M, A prospective cohort study. Atherosclerosis come of Mexican patients with first-ever
Deulofeu R, Aponte JH: Prognostic signifi- 2010;209:215–219. acute ischemic stroke: the PREMIER study.
cance of uric acid serum concentration in pa- 19 Dimitroula HV, Hatzitolios AI, Karvounis Rev Neurol 2010;51:641–649.
tients with acute ischemic stroke. Stroke 2002; HI: The role of uric acid in stroke: the issue 27 Chiquete E, Cantú-Brito C, Villarreal-Carea-
33:1048–1052. remains unresolved. Neurologist 2008; 14: ga J, Murillo-Bonilla LM, Rangel-Guerra R,
10 Scherbakov N, Sandek A, Martens-Lobenhof- 238–242. León-Jiménez C, et al: [Obesity paradox and
fer J, Kung T, Turhan G, Liman T, et al: En- 20 Weir CJ, Muir SW, Walters MR, Lees KR: Se- functional recovery in first-ever acute isch-
dothelial dysfunction of the peripheral vascu- rum urate as an independent predictor of emic stroke survivors: the PREMIER study].
lar bed in the acute phase after ischemic poor outcome and future vascular events after Rev Neurol 2010;51:705–713.
stroke. Cerebrovasc Dis 2012;33:37–46. acute stroke. Stroke 2003;34:1951–1956. 28 Cantú-Brito C, Ruiz-Sandoval JL, Murillo-
11 Romanos E, Planas AM, Amaro S, Chamorro 21 Karagiannis A, Mikhailidis DP, Tziomalos K, Bonilla LM, Chiquete E, León-Jiménez C,
A: Uric acid reduces brain damage and im- Sileli M, Savvatianos S, Kakafika A, et al: Se- Arauz A, et al: The first Mexican multicenter
proves the benefits of rt-PA in a rat model of rum uric acid as an independent predictor of register on ischaemic stroke (the PREMIER
thromboembolic stroke. J Cereb Blood Flow early death after acute stroke. Circ J 2007; 71: study): demographics, risk factors and out-
Metab 2007;27:14–20. 1120–1127. come. Int J Stroke 2011;6:93–94.
12 Hong JM, Bang OY, Chung CS, Joo IS, Gwag 22 Higgins P, Ferguson LD, Walters MR: Xan- 29 Amaro S, Urra X, Gómez-Choco M, Obach V,
BJ, Ovbiagele B: Influence of recanalization thine oxidase inhibition for the treatment of Cervera A, Vargas M, et al: Uric acid levels are
on uric acid patterns in acute ischemic stroke. stroke disease: a novel therapeutic approach. relevant in patients with stroke treated with
Cerebrovasc Dis 2010;29:431–439. Expert Rev Cardiovasc Ther 2011;9:399–401. thrombolysis. Stroke 2011; 42(1 suppl):S28–
13 Brouns R, Wauters A, Van De Vijver G, De Sur- 23 Amaro S, Soy D, Obach V, Cervera A, Planas S32.
geloose D, Sheorajpanday R, De Deyn PP: De- AM, Chamorro A: A pilot study of dual treat- 30 Zhang B, Gao C, Yang N, Zhang W, Song X,
crease in uric acid in acute ischemic stroke cor- ment with recombinant tissue plasminogen Yin J, et al: Is elevated SUA associated with a
relates with stroke severity, evolution and out- activator and uric acid in acute ischemic worse outcome in young Chinese patients
come. Clin Chem Lab Med 2010;48:383–390. stroke. Stroke 2007;38:2173–2175. with acute cerebral ischemic stroke? BMC
14 Schretlen DJ, Inscore AB, Vannorsdall TD, 24 Amaro S, Cánovas D, Castellanos M, Gállego Neurol 2010;10:82.
Kraut M, Pearlson GD, Gordon B, et al: Serum J, Martí-Fèbregas J, Segura T, Chamorro A: 31 Nardi K, Milia P: Letter by Nardi and Milia
uric acid and brain ischemia in normal elder- The URICO-ICTUS study, a phase 3 study of regarding article, ‘translational stroke re-
ly adults. Neurology 2007;69:1418–1423. combined treatment with uric acid and rtPA search of the combination of thrombolysis
15 Dawson J, Lees KR, Weir CJ, Quinn T, Ali M, administered intravenously in acute isch- and antioxidant therapy’. Stroke 2011;42:e547.
Hennerici MG, et al: Baseline serum urate and aemic stroke patients within the first 4.5 h of 32 Poss WB, Huecksteadt TP, Panus PC, Free-
90-day functional outcomes following acute onset of symptoms. Int J Stroke 2010; 5: 325– man BA, Hoidal JR: Regulation of xanthine
ischemic stroke. Cerebrovasc Dis 2009; 28: 328. dehydrogenase and xanthine oxidase activity
202–203. by hypoxia. Am J Physiol 1996; 270:L941–
16 Kanellis J, Johnson RJ: Editorial comment – L946.
Elevated uric acid and ischemic stroke: accu- 33 George J, Struthers AD: Role of urate, xan-
mulating evidence that it is injurious and not thine oxidase and the effects of allopurinol in
neuroprotective. Stroke 2003;34:1956–1957. vascular oxidative stress. Vasc Health Risk
Manag 2009;5:265–272.