Renal function and outcome of PTRA and stenting for atherosclerotic renal artery stenosis

Kidney International, Vol. 63 (2003), pp. 276 282 Renal function and outcome of PTRA and stenting for atherosclerotic renal artery stenosis FELIPE RAMOS, CAROL KOTLIAR, DANIEL ALVAREZ, HUGO BAGLIVO, PABLO RAFAELLE, HUGO LONDERO, RAMIRO SÁNCHEZ, and CHRISTOPHER S. WILCOX Instituto de Cardiologia y Cirugia Cardiovascular (ICYCC), Fundación Favaloro, Buenos Aires, Argentina, and Division of Nephrology and Hypertension and Center for Hypertension and Renal Disease Research, Georgetown University, Washington, D.C., USA Renal function and outcome of PTRA and stenting for atherothe after correction of baseline data to exclude the influence of sclerotic renal artery stenosis. expected regression to the mean. Background. Prior studies of percutaneous transluminal recriteria Conclusions. Patients with atherosclerotic RAS fulfilling strict nal artery angioplasty and stenting (PTRAS) for atherosclerotic of severity may have significant improvements in BP renal artery stenosis (RAS) have shown that renal function is one year after PTRAS but only modest in GFR. The initial improved in about 25%, stabilizes in about 40%, but worsens GFR may anticipate whether the benefits in the outcome will in about 25% of patients. The factors predicting benefit remain be in renal function enhancement (those with an initially de- controversial. We tested the hypothesis that the baseline glonormal pressed GFR) or in hypertension control (those with an initially merular filtration rate (GFR) predicts the changes in GFR and or mildly impaired GFR). blood pressure (BP) after PTRAS. Methods. Treated hypertensive patients with positive renal color-coded duplex Doppler velocimetry and clinical criteria Renal artery stenosis (RAS) causing renovascular diswere screened by arteriography. Patients (N 105) were included if they had an RAS 70%, a transluminal pressure gradient ease was considered a rare cause of secondary hyperten- 30 mm Hg and, they had more than 100 days of follow-up. sion. However, with improvements in techniques for GFR was calculated from the serum creatinine concentration screening and with the widespread use of vascular im- (S Cr ). Patients were divided by baseline GFR into subgroups aging, it is now recognized that 40 to 50% of patients with normal to mildly impaired (N 52) or moderately to with occlusive disease of the lower limb [1] and 15 to 30% severely impaired (N 53) initial GFR, according to a GFR 50 or 50 ml min of patients with coronary artery disease have identifiable respectively. All received PTRAS. Results. For the entire group, after a mean follow-up period RAS [2]. Reports of patients with established, or newly of 371 days, there were significant reductions in systolic and diagnosed, end-stage renal disease (ESRD) indicate a diastolic BP (before, 160 26/91 12 vs. after, 145 20/83 prevalence of RAS of 10 to 22% [3, 4]. Some 3 to 30% 10 mm Hg, respectively; mean SD; P 0.0001), and a modest of patients with atherosclerotic RAS show a progressive increase in the calculated GFR (before, 54 26 vs. after, 62 loss of renal mass over three to five years [5]. A number 28 ml min 1 ; mean SD; P 0.007). However, in the subgroup of patients with an initially lower GFR there was a significant suffer renal artery occlusion, which may prevent subse- increase in the calculated GFR (from 33.3 10 to 54 24 quent intervention and often heralds a sharp decline in ml min 1 ; mean SD; P 0.0001) despite no significant renal function. Percutaneous transluminal renal angiochange in BP (161 27/90 12 vs. 151 21/86 12; P plasty and stenting (PTRAS) provides a much less inva- NS). In contrast, in the subgroup with an initially higher GFR, sive technique than surgery, yet also has a high probabilthere were significant (P 0.0001) reductions in systolic BP (from 159 25 to 138 16 mm Hg) and diastolic BP (from ity of correction of the stenosis. These considerations 91 11 to 81 9 mm Hg), but no significant change in the have led clinicians to be more aggressive in this search calculated GFR (from 75 21 to 70.2 30 ml min 1 ; P for RAS. However, recent controlled clinical trials have NS). The significance of GFR variation in subgroups remained cast doubt on the efficacy of angioplasty for RAS and have documented a significant rate of complications. Improvements in blood pressure (BP) or glomerular filtra- Received for publication October 5, 2001 and in revised form August 28, 2002 Accepted for publication August 30, 2002 2003 by the International Society of Nephrology 276 Key words: renal vascular hypertension, ischemic nephropathy, hyper- tension, glomerular filtration rate, percutaneous transluminal renal artery stenting. tion rate (GFR) were not detected in the DRASTIC trial in hypertensive patients with RAS 50% who were randomized to PTRA, compared to medical therapy, over a one year period of follow-up [6]. Nevertheless, meta-analyses of the seven trials reporting surgical intervention and of four reporting intervention with PTRAS

Ramos et al: GFR and BP after stenting 277 indicates that 25 to 30% had an improvement in GFR, tion recommendations [10]. Reported values for each 45 to 50% had a stable GFR, whereas 20 to 25% had a subject are the average of three separate readings. decline in GFR at follow-up [7, 8]. The failure to show Renal function was assessed from measurements of statistically significant effects of intervention on GFR in serum creatinine concentration (S Cr ) and application of individual studies was due to a similar number of those the Cockroft and Gault formula [11] to estimate the who improved or who deteriorated [7]. Therefore, it glomerular filtration rate (GFR; ml min 1 ) according becomes critical to define subgroups of patients with to the S Cr (mg dl 1 ), weight (kg), age (years) and gender RAS who may have an improvement in BP or GFR after of the patient: intervention. We tested the hypothesis that, in selected patients with GFR (males) [(140 age) body weight]/(s Cr 72) quite severe RAS, the preexisting level of renal function GFR (females) Value for males 0.85 might predict a group of patients who will experience a favorable GFR and/or BP response to intervention with Renal arteriography was performed through a femoral PTRAS. artery access with a guided 8-French catheter placed in the aorta to study selectively all renal arteries. Thereafter, if a renal artery stenosis 70% was detected, further METHODS views were undertaken (in the anteroposterior and coronal The Ethical Review Board of the Favaloro Foundation planes) to confirm the degree of luminal stenosis in Buenos Aires, Argentina where all of these studies from the arteriographic photographs. The catheter was were conducted, authorized the study. Informed consent advanced across the stenosis and the pressure gradient was required in all participants. Patients selected were established, and in those cases with critical RAS the those at risk for RAS according to clinical criteria (JNC thinnest available catheter was employed. Whenever a VI stages II-III or resistant hypertension, or associated stenosis was considered hemodynamically significant renal failure or atherosclerotic vascular disease) and a ( 30 mm Hg pressure gradient), PTRA was performed positive screening test utilizing renal ultrasound with with a Tittan balloon where the width and length were color-coded duplex Doppler velocimetry of the renal selected according to the anatomy of the stenosis. There- arteries (HP SONOS 2500, with 2.5 and 3.5 MHz probes). after, a Palmaz Schatz stent was implanted over a Cordis Single operator echographic sensitivity and specificity balloon, which was inflated to 10 to 12 atmospheres or for repeated measurements for stenosis over 60% were as necessary to produce a good anatomical result [12]. both 97% (analysis of 343 arteries). There was a significases Complications observed after the procedure were two cant correlation between the quantitative estimations of of bleeding and large inguinal hematoma with an the degree of RAS by the duplex Doppler velocimetry hematocrit drop of more than 5%, and one case of right method with the values obtained subsequently by arteridure perirenal hematoma during a bilateral stenting proce- ography (r 0.79, P 0.001). From an initial sample managed conservatively without afterward clinical of 149 patients, a study group of 105 was obtained with consequences. significant atherosclerotic renovascular stenosis (detailed The following definitions were applied: later in this article), and a follow-up evaluation after at least 100 days. A number of 44 patients were excluded Hypertension: systolic blood pressure (SBP) 140 because they (a) had a RAS 70% in arteriography and mm Hg and/or diastolic blood pressure (DBP) 90 did not undergo PTRAS (N 27); (b) received angio- mm Hg, calculated as the average of three separate plasty without stenting (N 2); (c) were submitted to blood pressure recordings, or the need for current surgery (N 2); (d) had a primary stent failure (N 1); antihypertensive therapy. and (e) were lost to follow-up (N 12). The follow-up Significant renal artery stenosis: 70% luminal reduc- after PTRAS averaged 371 days (median 376, range 100 tion of a renal artery and a pressure gradient of 30 to 700). Only patients with primary technical success mm Hg across the lesion. following PTRAS were included [9]. Inclusion in the study Bilateral renal artery stenosis: 50% bilateral luminal required at least two clinic visits following the intervention. reductions. At variance with the former definition, Routine laboratory tests and a second renal ultra- a lower limit was used only in order to characterize sound study were undertaken in all study patients at the a subgroup of patients with higher extension of ath- last of these visits. Blood pressure data were the average erosclerotic renal artery disease suggesting a higher values obtained before the intervention and at the last risk of disease progression, but stenting were limited of the follow-up evaluations, both under prescribed anti- only to those defined as significant RAS. hypertensive treatments. A trained operator measured Cure of hypertension: SBP 140 mm Hg and DBP 90 BP with a mercury sphygmomanometer after two min- mm Hg in the absence of antihypertensive therapy. utes of sitting, according to the American Heart Associa- Renal function according to GFR: normal to mildly

278 Ramos et al: GFR and BP after stenting Table 1. Clinical and laboratory data before intervention Normal to mildly Moderate to severely Total impaired renal function impaired renal function population GFR 50 ml min 1 GFR 50 ml min 1 Number of patients 105 52 53 Age years 59 10 59 9 60 11 Gender female/male 42/63 21/31 21/32 SBP mm Hg 160 26 159 25 161 27 DBP mm Hg 91 12 91 11 90 12 Controlled hypertension BP 140/90 26 14 12 Number of antihypertensive drugs used 2.2 1.86 1 2.43 0.9 b Serum creatinine mg dl 1 1.67 0.88 1.03 0.26 2.29 0.83 Calculated (Cockroft-Gault) GFR ml min 1 54 26 75 20.6 33.6 9.6 Number with 3 risk factors a 68 33 35 Number with diabetes mellitus 29 15 14 Unilateral/bilateral renal artery lesions 60/45 34/18 26/27 Number with bilateral stents 18 7 11 Number with ostial stenosis 34 13 21 Abbreviations are: SBP, systolic blood pressure; DBP, diastolic blood pressure; GFR, glomerular filtration rate. a Hypertension, diabetes, dyslipidemia, tobacco use, obesity (BMI 27) b P 0.01, between subgroups using unpaired t test impaired 50 ml min 1, moderate to severely tensive and receiving drug treatment, but only 26 (25%) impaired 50 ml min 1. were controlled. Renal artery stenosis was found bilat- Primary technical success after PTRAS: residual stebilateral eral in 45 (43%) patients. Of these, 18 (40%) underwent nosis 50% and a pressure gradient 5 mm Hg withapplied PTRAS. A total number of 123 stents were out major complications (surgery required or death). and 33% of patients presented with ostial renal artery stenosis. Statistical analysis The BP and renal function at baseline before PTRAS Data are expressed as mean values standard deviaa and at follow-up is detailed in Table 2. For the group as tion. To assess within-patient changes, a paired t test was whole, there was a highly significant reduction in BP, applied, or Wilcoxon signed-rank test for paired nonquent a modest, but significant, reduction in S Cr, and a conse- normal distributed variables (serum creatinine). The population increase in the calculated GFR. sample was analyzed as a whole and by subgroups As shown in Figure 1, it is apparent from the sub- selected according to renal function as defined above. group s analysis that the outcome of PTRAS diverged Differences between subgroups were analyzed by unpaired upon the initial level of GFR. Thus, a significant improvet test. Univariate and multivariate regression analysis ment in BP control occurred only in the subgroup with and 2 tests were applied when appropriate (GB-STAT initially higher renal function. Conversely, a reduction version 7.0; Dynamic Microsystems, Inc., Silver Spring, in the S Cr and an improvement in the calculated GFR MD, USA). A further analysis of the variation observed occurred only in the subgroup with initially more impaired at follow-up in the calculated GFR was undertaken after renal function. After application of the correction for recorrection in baseline GFR for the expected regression gression toward the mean, the improvement in renal func- toward the mean (RTM). This analysis was required tion in the subgroup of patients with initially impaired because of the selection process of subgroups according renal function remained highly significant (Table 2). to baseline renal function [13]. For this purpose, the After PTRAS there was a considerable improvement Shepard and Finison formula was applied [14 16]. A rewith in hypertension control. Overall, the fraction of patients peatability correlation index for GFR of 0.7 was applied. normotension increased from 25% at baseline to Differences were considered significant at P 0.05. 65% after PTRAS, including 19 patients who were cured of hypertension. A reduction in the number or doses of RESULTS antihypertensive drugs prescribed was observed in 67% of the patients who still required treatment. The clinical characteristics of the study population at Univariate regression analysis demonstrated a signifibaseline are summarized in Table 1. No differences in cant negative correlation between the baseline GFR and the analyzed variables were found between subgroups, the follow-up values for SBP (r 0.29, P 0.0026) with the exception of the number of prescribed antihy- and DBP (r 0.33, P 0.0006). After a multivariate pertensive drugs, which were higher in the group with stepwise regression analysis including age, baseline systolic more impaired renal function. All patients were hyper- (SBP) and diastolic (DBP) blood pressure, and

Ramos et al: GFR and BP after stenting 279 Table 2. Blood pressure and renal function before intervention and at follow-up Paired t test or Wilcoxon (S Cr ) Baseline data Follow-up data P value Total population (N 105) Systolic blood pressure mm Hg 160 26 145 20 0.0001 Diastolic blood pressure mm Hg 91 12 83 10 0.0001 Serum creatinine mg dl 1 1.67 0.88 1.38 0.68 0.009 Calculated (Cockroft-Gault) GFR ml min 1 54 26 62 28 0.007 Subgroup with normal to mildly impaired renal function (GFR 50 ml min 1, N 52) Systolic blood pressure mm Hg 159 25 138 16 0.0001 Diastolic blood pressure mm Hg 91 11 81 9 0.0001 Serum creatinine mg dl 1 1.03 0.27 1.21 0.54 NS Calculated (Cockroft-Gault) GFR ml min 1 75 26 70.2 30 NS GFR (basal data) after correction for RTM ml min 1 68.7 21 70.2 30 NS Subgroup with moderate to severely impaired renal function (GFR 50 ml min 1, N 50) Systolic blood pressure mm Hg 161 27 151 21 NS Diastolic blood pressure mm Hg 90 12 86 12 NS Serum creatinine mg dl 1 2.29 0.83 1.55 0.75 0.0001 Calculated (Cockroft-Gault) GFR ml min 1 33.3 9.6 54 24.1 0.0001 GFR (basal data) after correction for RTM a ml min 1 39.25 9.6 54 24.1 0.001 RTM is regression toward the mean. a Corrected for bias due to RTM by the Shepard and Finison formula [14]. GFR at follow-up as covariates, the relationship between Those with an initially elevated S Cr had a significant improvement SBP and DBP at follow-up as dependent variables with in their calculated GFR, despite no apparent baseline GFR remained significant. This indicates that benefit in their BP (BP decreased, though non-signifi- a low GFR value at baseline emerged as a negative cantly). In contrast, those with an initially normal S Cr had a predictor of PTRAS to reduce the systolic and diastolic significant improvement in their BP despite no apparent blood pressure at follow-up. changes in their calculated GFR. As anticipated, the re- A post-hoc subgroup analysis showed that bilateral stenosis rate was higher in patients with ostial stenosis, renal artery stenosis, which was detected in 45 patients, but remained low (14%) after one year of follow-up. apparently did not influence the changes in blood presthe The subgroups did not differ in their demographics, sure or GFR after intervention, although this conclusion levels of systolic and diastolic BP, the presence of is based on a limited number of patients. In 93% of comorbid conditions, risk factors, bilateral RAS, the rate patients, renal artery patency after stenting was evalu- of bilateral stenting or the fraction with ostial lesions. ated at follow-up using color duplex Doppler velocimetheir By design, they had significant differences at baseline in try. A re-stenosis was detected in 14 patients (14%). This levels of renal function. We evaluated whether was more likely in those with ostial lesions (27 vs. 8%; these results could be ascribed to the phenomenon of P 0.02 by Yates-corrected 2 ). regression toward the mean (RTM), since bias will occur when a population is divided into groups according to DISCUSSION one variable, and measurements are made sequentially with that variable as the outcome measure. In this cir- This study was conducted in 105 patients, who all under- cumstance, even in the absence of a real change, the went a PTRAS after applying quite strict criteria for RAS data for the two groups will converge [13 16]. Thus, requiring intervention. They were screened by clinical subsequent trends in the data for each group could be criteria and by positive renal artery color-coded duplex the effects of selection. The degree of RTM depends on Doppler velocimetry test. Entry required arteriographic the spontaneous variability of the parameter over time findings of anatomical RAS 70% and hemodynamic and on the differences between the individual mean valsignificance as evidenced by a gradient in arterial pres- ues of the subgroups from the whole population [13, 16]. sure across the stenosis 30 mm Hg. Blood pressure and A statistical correction was made in the baseline calcu- S Cr were evaluated before and at an average of one year lated GFR to compensate for the effects of RTM. Using after the intervention. The group as a whole had a sig- this correction, the modest increase in the S Cr and denificant improvement in their BP and calculated GFR, crease in the calculated GFR after intervention in the without important clinical complications or individuals subgroup of patients with initially normal renal function with renal function deterioration. However, the main was judged not statistically significant, while the more new findings were disclosed in the subgroup analysis. robust improvements in these variables in the subgroup

280 Ramos et al: GFR and BP after stenting Fig. 1. Blood pressure (BP) and glomerular filtration rate (GFR) changes after percutaneous transluminal renal arterial angioplasty and stenting (PTRAS). Symbols are: ( ) subgroup with GFR 50 ml min 1 ;( ) subgroup with GFR 50 ml min 1 ;( ) baseline GFR corrected for RTM; **P 0.001, ***P 0.0001. of patients with initially impaired renal function was maintained at a very high level of statistical significance. This study has some limitations. First, the GFR was calculated from the Cockcroft and Gault formula rather than being measured directly by a clearance technique. However, this method obviates errors due to incomplete urine collection, and repeated measurements in the same individual (as in this study), minimize errors due to changes in creatinine production [17]. Second, we screened for re-stenosis by Doppler rather than by the invasive method of arteriography. We selected Doppler because it is very sensitive for detecting re-stenosis after PTRAS [18]. Indeed, we found that the degree of RAS estimated by Doppler correlated closely with that measured di- rectly by arteriography. The rates of re-stenosis after PTRAS are only 25% of those for PTRA alone [19]. Third, the study lacked a control (untreated) group. Nev- ertheless, previous work has shown that atherosclerotic RAS tends to be progressive [7, 20]. In a recent study, 16% of patients with RAS who received only medical therapy progressed to complete occlusion of the affected renal artery over one year [5]. Therefore, the overall gain in the calculated GFR over one year can reasonably be ascribed to the intervention, rather than to a spontaneous improvement. Last, the study conclusions are valid only for the one-year period during which this group of patients was observed, and according to a non-standardized antihypertensive treatment. Long-term morbidity and mortality are important objectives that were not addressed in this study. An apparent recent increase in the prevalence of RAS [7, 20 23] has coincided with a more widespread availability of PTRA(S). Even the casual finding of RAS at the time of coronary or ilio-femoral arteriography can prompt a PTRA(S) in the hope of preserving renal function or improving control of hypertension. However, recent controlled clinical trials cast doubt on the wisdom of this practice [24]. The Scottish and Newcastle Renal Artery Stenosis Collaborative Group randomized 55 patients [25] and the EMMA trial randomized 49 patients [26] with RAS to PTRA or aggressive medical therapy. After six months of follow-up, there were no significant overall advantages of intervention on BP in either trial. The DRASTIC study detected no differences in BP or GFR between patients with RAS followed up for one year after randomization to PTRA or medical therapy [6]. One reason for the different conclusion from our study compared to controlled trials is that PTRAS has better results than PTRA alone [9, 12, 19, 21 23, 27 32]. In the present study, the rate of re-stenosis at an average follow- up of one year was 14%. A second reason that may account for differences between conclusion from this study and the controlled trials is the selection of patients. The DRASTIC study required only moderate or severe hypertension and a 50% RAS. This will include many patients at an early stage of RAS that is not yet functionally significant, diluting the beneficial effects of PTRA in the overall results. In contrast, we used a combination of an anatomical criterion of a stenosis of 70% and a functional criterion of an arterial pressure gradient 30 mm Hg across the stenosis, to eliminate patients with less severe RAS who may not derive benefit from the procedure. These inclu- sion criteria might help select the patients that will bene- fit from PTRAS. On the other hand, as patients not fulfilling the inclusion criteria were not submitted to PTRAS, the possible benefit in these subjects cannot be excluded. Another study should be aimed to answer this important point, that is, the relationship of the severity of vascular disease and the subsequent improvement obtained after PTRAS. To answer this question, the variables to be assessed should be thresholds of anatomic RAS, trans-stenotic BP gradients, levels of baseline GFR,

Ramos et al: GFR and BP after stenting 281 or other markers that could be useful to predict patient with normal renal function studied by split renal function outcome and to be applied in decision-making. methods, an improvement in GFR of the post-stenotic Several other studies have evaluated renal function kidney after angioplasty is offset by a decline in GFR after intervention for atherosclerotic RAS [7, 12, 27, 28]. of the contralateral kidney that limits the beneficial ef- Among the larger trials, Jensen et al studied 107 patients fects on overall GFR [35]. Discordant effects on GFR with atherosclerotic RAS over one year following PTRA and BP in the group with an initially elevated S Cr may [33]. There were significant (P 0.001) reductions in be related to the persistence of residual renal dysfunction, the BP and increases in the GFR. The GFR was measured since the GFR was not normalized after PTRAS. from the plasma clearance of ( 51 Cr) ethylenediamine- Intrarenal vascular disease predicts a poor BP response tetraacetic acid (EDTA) and increased from 48 24 to PTRA [12]. Despite the observed BP reduction was ml min 1 to 53 27 ml min 1. This is very close to not significant for the subgroup with more severe impairment the increase in calculated GFR from 54 to 62 ml min 1 in renal function, considering the observed reducthe detected in the present series. Watson et al studied renal tion in the drug s number/dose, a response of lesser magnitude function from the slope of the reciprocal of S Cr over time but of clinical value cannot be excluded, although in 33 azotemic patients with atherosclerotic RAS before, to confirm this effect a higher number of patients would and for 20 months after PTRAS [34]. Renal function, be required. In the trials comparing the effects on BP which was deteriorating before the intervention, was sta- after PTRAS, more patients were cured of hypertension bilized thereafter. Similar results were reported in a in the group with an initially normal S Cr value [28], as in group of 23 patients with RAS whose deteriorating renal our study. Although renal dysfunction has been considered function was stabilized by PTRAS [23]. Our results extend a marker of adverse outcomes after PTRAS [32], these findings by demonstrating that beneficial ef- our results showing a significant improvement in GFR fects on renal function may be anticipated in selected in patients with RAS and renal function impairment in patients treated with PTRAS if their baseline renal func- a sample of subjects younger than other reported series, tion is abnormal. Whether this result may imply an overall raise questions about the prognostic role of other markprognostic improvement in prognosis, based on the well-known ers like hypertension control (worse in those with pre- value of serum creatinine, or conversely, as viously altered GFR), age, and comorbidity. Studies de- serum creatinine was not entirely normalized, the rela- signed with hard end points, prolonged follow-up and tive prognostic value of the baseline creatinine may be controlled interventions over the main treatable risk fac- more important than serum creatinine at outcome after tors (such as hypertension, dyslipidemia, diabetes) are PTRAS, remains a matter of speculation, but this issue needed to find these answers deserves further study. Dorros et al did not address that point, and they underscored the importance of PTRAS Conclusions as a revascularization procedure before the onset of renal Two conclusions can be drawn from our study that, if dysfunction in order to improve BP control, preserve or confirmed, may help select patients most likely to benefit prevent renal function deterioration and patient survival from intervention. First, PTRAS can reduce BP and increase [32]. Based in our observations, the improvement of GFR over a mean of one year in a group of GFR after PTRAS in patients with decreased renal func- patients carefully selected to have high grade anatomical tion resulting in a partial restoration or an interruption RAS that is hemodynamically significant. Second, among of the downward slope of renal function along the spon- such selected patients, those with initially normal to taneous evolution of ischemic nephropathy as reported mildly impaired renal function may anticipate an im- by Harden et al [23] and Watson et al [34] supports a proved BP after intervention, whereas those with initially valuable therapeutic role for PTRAS in such patients, moderate to severely impaired renal function may antici- possibly delaying the onset of ESRD. pate an improved GFR. This latter conclusion remains An interesting finding was the discordance between strongly supported after the correction for the RTM the beneficial effects of PTRAS on BP and GFR in the phenomenon in baseline GFR data. analysis of subgroups. The group with an initially higher GFR experienced a beneficial effect on BP without a ACKNOWLEDGMENTS change in GFR, whereas the group with an initially lower Work in Christopher S. Wilcox s program is supported by funds GFR experienced a beneficial effect on GFR but not a from the George E. Schreiner Chair of Nephrology. A portion of this significant change in BP. With methods based on S study was presented as a poster at the 2000 ISH meeting (Chicago). Cr We thank Ms. Sharon Clements for expert presentation of the manudeterminations it is not possible to detect a small change script and Mr. Carlos A. Giannone for statistical advice and analysis. in the calculated GFR within the normal range. Therefore, the failure to see a change in the calculated GFR Reprint requests to Felipe Ramos, M.D., Sección Hipertensión Arte- rial, Instituto de Cardiología y Cirugía Cardiovascular, Fundación Favain the group with initially normal renal function may be loro, Belgrano 1746, 1039, Buenos Aires, Argentina. a limitation of the technique used. Moreover, in patients E-mail: framos@ffavaloro.org

282 Ramos et al: GFR and BP after stenting REFERENCES phy in assessing restenosis of renal artery stents. Cardiovasc Intervent Radiol 22:475 480, 1999 1. Choudri AH, Cleland JGF, Rowlands PL, et al: Unsuspected 19. Van de Ven PJG, Beutler JJ, Kaatee FJA, et al: Transluminal renal artery stenosis in peripheral vascular disease. BMJ 301:1197 vascular stent for ostial atherosclerotic renal artery stenosis. Lancet 1198, 1990 346:672 674, 1995 2. Uzu T, Inoue T, Fujii T, et al: Prevalence and predictors of renal 20. Textor SC: Epidemiology and clinical presentation. Semin Nephrol artery stenosis in patients with myocardial infarction. Am J Kidney 20:426 431, 2000 Dis 29:733 738, 2000 21. Van de Ven PJG, Kaatee R, Beutler JJ, et al: Arterial stenting 3. Mailloux LU, Napolitano B, Bellucci AG, et al: Renal vascular and balloon angioplasty in ostial atherosclerotic renovascular disdisease causing end-stage renal disease, incidence, clinical corre- ease: A randomized trial. Lancet 353:282 286, 1999 lates, and outcomes: A 20-year clinical experience. Am J Kidney 22. Bloch MJ, Trost DW, Pickering TG, et al: Prevention of recurrent Dis 24:622 629, 1994 pulmonary edema in patients with bilateral renovascular disease 4. Beutler JJ, Van Ampting JMA, Van de Ven PJG, et al: Long- through renal artery stent placement. Am J Hypertens 12:1 7, 1999 term effects of arterial stenting on kidney function for patients with 23. Harden PN, MacLeod MJ, Rodger RSC, et al: Effect of renal- ostial atherosclerotic renal artery stenosis and renal insufficiency. J artery stenting on progression of renovascular renal failure. Lancet Am Soc Nephrol 12:1475 1481, 2001 349:1133 1136, 1997 5. Caps MT, Zierler RE, Polissar NL, et al: Risk of atrophy in 24. Textor SC, Wilcox CS: Ischemic nephropathy or azotemic renovascular disease. Semin Nephrol 20:489 502, 2000 kidneys with artherosclerotic renal artery stenosis. Kidney Int 25. Webster JF, Marshall M, Abdalla A, et al: Randomised compar- 53:735 742, 1998 ison of percutaneous angioplasty vs. continued medical therapy 6. Van Jaarsveld B, Krijnen P, Pieterman H, et al: The effect of hypertensive patients with atheromatous renal artery stenosis. balloon angioplasty on hypertension in atherosclerotic renal-artery Scottish and Newcastle Renal Artery Stenosis Collaborative Group. stenosis. N Engl J Med 342:1007 1014, 2000 J Hum Hypertens 12:329 335, 1998 7. Textor SC, Wilcox CS: Renal artery stenosis: A common, treat- 26. Plouin PF, Chatellier G, Darne B, Raynaud A: Blood pressure able cause of renal failure? Annu Rev Med 52:421 442, 2001 outcome of angioplasty in atherosclerotic renal artery stenosis: A 8. Textor SC: Revascularization in atherosclerotic renal artery dis- randomized trial. Essai Multicentrique Medicaments vs Angioplasease. Kidney Int 53:799 811, 1998 tie (EMMA) Study Group. Hypertension 31:823 829, 1998 9. Leertouwer TC, Gussenhoven EJ, Bosch JL, et al: Stent place- 27. Bloch MJ, Pickering TG: Renal vascular disease: Medical manment for renal arterial stenosis: Where do we stand? Radiology agement, angioplasty and stenting. Semin Nephrol 20:474 488, 2000 216:78 85, 2000 28. Isles CG, Robertson S, Hill D: Management of renovascular 10. Frohlich ED, Grim C, Labarthe DR: Recommendations for hu- disease: A review of renal artery stenting in ten studies. Q J Med man blood pressure determination by sphygmomanometers: Re- 92:159 167, 1999 port of a special task force appointed by the Steering Committee, 29. Blum U, Krumme B, Flugel P, et al: Treatment of ostial renal- American Heart Association. Hypertension 11:209A 222A, 1988 artery stenoses with vascular endoprostheses after unsuccessful 11. Cockcroft DW, Gault MH: Prediction of creatinine clearance balloon angioplasty. N Engl J Med 336:459 465, 1997 from serum creatinine. Nephron 16:31 41, 1976 30. White CJ, Ramee SR, Collins TJ, et al: Renal artery stent place- 12. Rees CR: Stents for atherosclerotic renovascular disease. J Vasc ment: Utility in lesions difficult to treat with balloon angioplasty. Interv Radiol 10:689 705, 1999 J Am Coll Cardiol 30:1445 1450, 1997 13. Ingelfinger JA, Mosteller F, Thibodeau LA, Ware JH: Biostasclerotic ostial renal artery stenosis with the intravascular stent. 31. Tuttle KR, Chouinard RF, Webber JT, et al: Treatment of atherotistics in Clinical Medicine (3 rd ed). New York, McGraw-Hill, Inc., Am J Kidney Dis 32:611 622, 1998 1994 32. Dorros G, Jaff M, Mathiak L, et al: Four-year follow-up of Pal- 14. Shepard DS, Finison LJ: Blood pressure reductions: Correcting maz-schatz stent revascularization as treatment for atherosclerotic for regression towards the mean. Prev Med 12:304 317, 1983 renal artery stenosis. Circulation 98:642 647, 1998 15. Altman DG, Bland JM: Analysis of blood pressure data (chapt 33. Jensen G, Zachrisson BF, Delin K, et al: Treatment of renovascu- 15), in Handbook of Hypertension (vol 14), edited by O Brien E, lar hypertension: One year results of renal angioplasty. Kidney Int O Malley K, Amsterdam, Elsevier, 1991, pp 306 311 48:1936 1945, 1995 16. Yudkin PL, Stratton IM: How to deal with regression to the 34. Watson PS, Hadjipetrou P, Cox SV, et al: Effect of renal artery mean in intervention studies. Lancet 347:241 243, 1996 stenting on renal function and size in patients with atherosclerotic 17. Agodoa L, Eknoyan G, Ingelfinger J, et al: Assessment of struc- renovascular disease. Circulation 102:1671 1677, 2000 ture and function in progressive renal disease. Kidney Int 52(Suppl 35. Batide-Alanore LA, Azizi M, Froissart M, et al: Split renal 63):S144 S150, 1997 functions outcome after renal angioplasty in patients with unilateral 18. Bakker J, Beutler JJ, Elgersma OEH, et al: Duplex ultrasonogra- renal artery stenosis. J Am Soc Nephrol 12:1235 1241, 2001