Renal Function and Survival After Renal Artery Stent Revascularization May Be Influenced by Embolic Debris

ABSTRACT: Four years of follow-up renal function and survival data were obtained on 544 patients who underwent 714 successful renal artery stenosis (RAS) stent revascularizations. The mean serum creatinine (SCr) was unchanged at 4 years (1.6 ± 1.0 mg/dl versus 1.6 ± 0.9 mg/dl). The 2-year paired comparison analysis revealed no change in the reciprocal of the SCr (1/SCr), when compared with baseline or 1-year follow-up values. Simple linear regression analysis revealed flat post-procedure trend line slopes within all patient subsets, which implied renal functional stabilization. The 4-year mortality rate was 20%; variables directly affecting survival were presence of 1 or 2 kidneys, whether 1 or 2 renal arteries were stenotic [survival: unilateral 87%, bilateral 80%, and solitary kidney 60% (p J INVAS CARDIOL 2004;16:189–195 Key words: renal artery stenosis, revascularization, stent Atherosclerotic renal artery stenosis (RAS) has resulted in accelerated or poorly controlled hypertension, renal dysfunction, pulmonary edema,^1–8 ischemic atrophy⁹ and ischemic nephropathy, of which histopathologic severity correlated with renal function.¹⁰ RAS stent revascularization has resulted in more facile blood pressure control^11–30 and preservation or stabilization of renal function.^24–33 This communication details the 4-year renal function follow-up and survival data of 544 RAS patients who underwent successful stent revascularization, utilizing actuarial, paired comparison and simple linear regression analyses. Methods Between 1990 and 1997, five hundred forty-four patients underwent balloon-expandable stent revascularization (Palmaz™, Palmaz-Schatz™; Cordis Corporation, Warren, New Jersey) of 714 atherosclerotic stenotic renal arteries at two sites. Baseline and follow-up data were prospectively collected and collated on successful patients; procedural methodologies, and consent forms were reviewed and approved by each Investigational Review Board. No significant differences existed between the cohort demographics of the two sites. Patient inclusion and exclusion criteria, renal function assessment and group stratification design have been reported.^29,30Statistical analysis. The 2-year paired comparison analyses compared data at 6, 12 and 24 months with the patients’ baseline values. The 4-year simple linear regression analyses assessed renal function with a trend line slope. The Kaplan-Meier life table methodology assessed survival. A p-value of Renal data: Function. Paired comparison analysis showed that the 1/SCr slope was flat, which inferred functional stabilization. If ischemic nephropathy were a continuous, progressive disease, then an alteration in functional assessment would imply a change in renal physiology. Both case studies demonstrated functional improvement, which would not be consistent with the dogma that ischemic functional deterioration was an ongoing, irreversible progressive process, apart from a few conditions.^37–39 However, this premise of progressive functionally worsening has been challenged. Harden³⁰ demonstrated that 69% of 32 renal stent patients had improvement or stabilization of renal function, and that functional deterioration had been significantly slowed [the 1/SCr’s slope changed from -4.34 to -0.55 L.mumol/day (p 40 showed that the negative 1/SCr slope of 33 chronic renal insufficiency patients, with 61 stent revascularized vessels, became positive [-0.0079 to 0.0043 dl.mg/month (p 34 concluded that if renal function was stable prior to the procedure, then SCr levels would not change. However, if renal function was declining, then improvement occurred during the first year [SCr fell from 182 µmol/L to 154 µmol/L (p 9 continued pole-to-pole kidney length shortening in non-revascularized RAS kidneys. Tuttle’s³² assessment of the creatinine clearance and glomerular filtration rate (GFR), utilizing Cockroft-Gault methodology, found that the SCr remained stable in his 129 renal stent patients, of whom 57% had dysfunction. However, a transitory functional impairment was detailed. Cohort’s baseline filtration rate was 40 ± 2 ml/minute, and subsequent values were 36 ± 3 ml/minute, 39 ± 3 ml/minute and 39 ± 4 ml/minute at 6, 12 and 24 months, respectively. Patients with severe dysfunction (SCr >= 2.0 mg/dl) showed a similar 6-month trough: baseline of 53 ± 3 ml/minute and 43 ± 4 ml/minute, 46 ± 4 ml/minute and 52 ± 5 ml/minute at 6, 12 and 24 months, respectively. Thus, an initial diminution in creatinine clearance appeared to have self-corrected. Simple linear regression analysis data, using 1/SCr, would not reveal the transitory change seen with GFR measurements; however, the functional stabilization reflected by the 1/SCr slope implied stabilization of the GFR, as well as a concordance between the SCr and GFR. Renal data: Concordance of the glomerular filtration rate and serum creatinine. Creatinine clearance is the summation of creatinine creation, minus its renal and extrarenal elimination. The aforementioned data implied that the GFR⁴⁴ and SCr were congruent, and thus that the SCr was a good functional marker. The SCr has been a simple method of functional assessment, and, by corollary, an outcome assay of any renal intervention. However, the SCr does not perfectly reflect the GFR,^42,43 which is the summation of the filtering of all functioning nephrons. At normal levels of renal function (SCr = 1.5 mg/dl), the SCr became more concordant with the GFR, and, as such, with function. This disparity between the SCr and GFR in normal function patients makes the reliability of SCr measurements questionable in normal patients. First, glomerular pathology may be veiled by a normal SCr. If the relatively uniform glomerular pathology of ischemic nephropathy were present, then the total GFR, reflecting a uniform diminution in glomerular capillary surface and/or permeability, would produce a concordant elevation in the SCr after sufficient GFR impairment. However, if nephrons were not all equally or uniformly affected, then the total GFR, which reflected the glomerular capillary surface and/or membrane permeability of both healthy and pathologically functioning glomeruli, could remain within the normal range if functioning glomeruli could overcome the deficit incurred by the pathologically minimally effected or non-functioning glomeruli. In fact, the total GFR and SCr, despite advanced pathological functional disturbances, could remain unchanged if sufficient augmented glomerular perfusion of non-effected or minimally effected glomeruli occurred. Thus, nephron impairment, uniformly diffuse or segmented, could exist despite a normal SCr. This premise has been supported by the wide range for normal GFR measurements (30 and 80 ml/min/1.73m2), while the GFR variation in dysfunctional kidneys was narrow. Thus, the functional stabilization, using SCr assessment in a dysfunctional cohort post-stent placement, could indicate that improved blood perfusion to these glomeruli, by relief of the renal artery obstruction, produced improved or augmented glomerular function, which resulted in a stabilized GFR, no further functional deterioration and, as a result, a stable SCr. La Batide-Alanore⁴⁴ performed split renal function studies on angioplasty patients with two kidneys and one stenotic renal artery. Single kidney GFR measurements, using Inulin or 51Cr-EDTA clearance at baseline and at 6 months showed that the “total GFR increased slightly significantly in the 29 patients with positive lateralization indices, the split renal function and single-kidney GFR of the stenotic kidney increased, whereas concurrently the GFR and split renal function of the non-stenotic kidney decreased significantly.” At 6 months, “a reversal of both hypoperfusion of the stenotic side and hyperperfusion of the non-stenotic side was observed, which was accompanied by a slight increase in total GFR.” Thus, in unilateral disease patients, glomerular filtration in the non-affected kidney, because of augmentation or compensation, maintained a normal SCr, and renal function improved following angioplasty. However, this explanation remains applicable only to two-kidney, single RAS patients, and not to bilateral disease/solitary kidney patients. In the latter two cohorts, the pathologic nephrologic process must be segmental or non-uniform, since no non-affected kidney existed, and as such no augmentation in the non-affected kidney could have occurred. Thus, if functional stabilization occurred, some glomeruli in these diseased kidneys must have functioned better after increasing renal perfusion. Intuitively, the pathology associated with chronic renal ischemia should and has caused uniform damage to each glomerulus. Tuttle’s transitory GFR changes provide possible evidence for the presence of a non-uniform or uneven pathology, which had arisen from recurrent atheroemboli, either from the renal artery stenosis or another site, e.g., atherosclerotic degenerated aorta. Other observations^{24,30,33,42,48} have similarly detailed this transient functional change, and some have detailed that atheroembolic showering occurred after stent placement. If atheroembolic debris released intermittently from an arterial stenosis lodged non-uniformly within the mesangium’s end-arteriole glomeruli, and if these arteriole blockages resulted in a segmental pathology, then progressive functional deterioration would result as more glomeruli became involved, but would not be apparent if augmentation occurred in non-affected glomeruli. As such, one aim of any RAS intervention would be the adjunctive prevention of additional emboli release, which could damage more and more glomeruli, especially in severely impaired patients with few functioning glomeruli. Therefore, ischemia may not be the primary cause of kidney dysfunction from RAS, since circumstantial evidence implicates recurrent atheroemboli as the major factor causing the non-uniform glomerular pathology and the subsequent renal dysfunction. In fact, the post-procedure stable renal function may reflect increased glomerular perfusion to those capable of functional augmentation. This physiologic change, as well as prevention of subsequent emboli episodes by a neoendothelial covered stent, resulted in stable function. Perhaps, if debris showers could be prevented during stent deployment, especially in patients with marginal glomerular function, then fewer glomeruli would be damaged and the transient decline in the GFR might not occur, which might be very important in dysfunctional patients. Furthermore, the hypothesis that the increased renal perfusion pressure, coupled with glomerular functional augmentation, resulted in stabilized or even improved function, was observed in the two case studies. Survival. The better survival of poorly controlled in contrast to medication controlled blood pressure patients is counterintuitive and contrary to accepted medical theory. Baseline data for the poorly controlled versus the controlled blood pressure patients detailed a lower SCr (1.7 ± 1.0 mg/dl versus 2.1 ± 1.8 mg/dl; p Study limitations. This study’s primary limitation is lack of more complete follow-up data, which has been a problem for a voluntary registry and also in clinical settings in which patients are treated by tertiary care interventionists who do not have primary control of the patient. In addition, once the patient has returned to the primary care physician, data often will not be collected at the appropriate times. However, the cohort size and the use of paired comparison and single linear regression analyses make these data compelling. Finally, while the study coordinator’s attempt to retrospectively obtain adequate pre-procedure serum creatinine data was unsuccessful, and as such precluded demonstration of an alteration in the post-stent revascularization, the case studies’ 1/SCr slope did demonstrate this. Conclusion. The follow-up of successfully stent revascularized RAS patients demonstrated renal function stability, which underscored the premise that the nephropathy associated with RAS may be primarily segmental and, more likely, a result of recurrent atheroemboli lodging within the end-arteriole glomerulus rather than from ischemia. The better survival of the poorly controlled blood pressure cohort with its lower baseline SCr than that of the medication controlled blood pressure cohort indicated that two pathologic processes probably existed, and that recurrent embolization rather than ischemia appeared to more adversely affect function and survival. Furthermore, if the shower of embolic debris could be eliminated during stent placement, this approach could not only decrease the number of immediately damaged glomeruli, but also, coupled with the neoendothelialization of the stent, might preclude further recurrent embolic damage and permit glomerular functional augmentation. The result of this approach might result in functional stabilization or even actual functional improvement, which could positively impact survival. Thus, the early diagnosis of RAS, especially in patients with medically managed hypertension, could avoid specious contentment achieved by medically controlled blood pressure, which unsuspectingly permitted unabated recurrent embolization and resulting renal dysfunction. A randomized trial of stent-supported RAS patients, with and without embolic protection devices, and periodic SCr measurements, especially in patients with renal dysfunction, could easily provide concrete answers to these clinical interrogatories. In addition, obtaining SCr values from dates prior to the stent procedure would enable demonstration of a change in the trend line slope, and whether progressive functional worsening could be halted. Positive conclusions of such a trial would have enormous medical, economic and social implications.