Renal Arterial Ultrasound – Predictor of Things to Come?
Surgical correction of renal arterial disease has been reported in the United States since 1975 and was primarily used in patients suffering severe or refractory hypertension or incipient renal failure with demonstrable critical renal artery stenoses. This method has largely been discarded in favor of percutaneous intervention, initially with balloon angioplasty alone and more recently with stent implantation. Despite the enthusiasm for correction of the anatomic lesion thought to be responsible for hypertension and loss of renal parenchymal function, many patients do not reap the intended benefit, either due to a misunderstanding of the physiologic significance of the treated lesion, progression of intrinsic renal disease, continued aortic atheroembolism or complications of the procedure itself. Because of these findings, renal artery revascularization for atherosclerotic renal arterial stenosis remains controversial, with a recent publication in the New England Journal of the ASTRAL study concluding that revascularization offers "no worthwhile clinical benefit" over medical therapy.
Of the feared sequelae of renal revascularization, embolization has been most often cited as a reason for continued renal tissue loss and decline of function. Small studies have suggested that some degree of embolization is virtually ubiquitous during renal stent implantations, and these findings are corroborated by the presence of emboli in debris aspirated from balloon-protected renal stents as well as trapped in filter protected renal stents. Suppose we were able to predict on the basis of pre-stent imaging which lesions were most likely to result in liberation of debris and worsening of renal function? A tantalizing report in this issue of the Journal by Prasad et al proposes to do just that. Using intravascular ultrasound "virtual histology," a method utilizing radiofrequency backscatter to identify the contents of arterial plaque, the authors relate the presence of necrotic core with a worsening of renal arterial flow after renal artery stenting as judged by renal frame count – the number of frames required for radiocontrast to reach the end arteries of the kidney. An increase in frame count means that contrast takes longer to fill the renal bed and is a harbinger of a poor result after renal intervention.
Using a clinically available intravascular ultrasound (IVUS) system with a phased array catheter and a virtual histology (VH) software package, the authors measured the plaque constituents, imaging throughout a vessel length beginning 5mm distal to the stented segment by continuous motorized pullback to the ostium of the renal artery. VH representations of fibrous tissue, fibrofatty tissue, necrotic core and dense calcifications were volumetrically quantified throughout the assessed vessel length and at the point of minimal luminal diameter within the lesion. The clinical features of the patients were tabulated, and interestingly, similarly to other studies of interventional patients, only 59% of those studied had been treated with HMG-CoA reductase inhibitors at the time of intervention. Patients were accessed using the no-touch technique, treated with routine predilatation, which is not agreed standard practice, and then stented. IVUS was performed prior to predilatation in 17 of the 18 patients. One lesion was so tight that it was judged to require predilatation prior to passing the stenosis with the IVUS catheter. One lesion was treated without stenting – we are not told why. Two lesions were not interrogated using continuous pullback, and only the minimal luminal diameter frame was analyzed in these two patients. No protection was used in this cohort.
Using this technique, the renal artery stenoses were reduced from a mean of 66% to 8.6%, and renal frame count for the overall group was reduced from 34.5 to 30.5 frames (mean reduction 2.5 ± 13 frames). 10 out of the 17 patients showed improvement in flow with a mean reduction in renal frame count of 11.9 ± 7.6 frames. The other 7 patients showed worsening of flow with a mean increase of 7.1 ± 8.0 frames.
There were two important findings of this study. First, analysis of the IVUS determined virtual histology showed that the stented segments were primarily composed of fibrous tissue (62.1%), with less necrotic core (15.6%), dense calcification (13.2%) and fibrofatty tissue (8.7%). Second, and most important, there was a clear relationship seen between an increase in the amount of necrotic core and an increase in renal frame count after stenting. In other words, the more necrotic core seen in the stented segment, the greater the increase in renal frame count, and thus the greater the loss of renal flow.
This was a small group of patients, only 18, and although a reduction in systolic blood pressure was seen over the whole group at 10.6 weeks in all patients (141 ± 22 mmHg baseline vs. 124 ± 8.3 mmHg at follow-up) no statements could be made about the relationship of plaque composition or renal frame count and the clinical outcomes of blood pressure control or change in GFR.
Although this was a small study, and therefore limited in the scope of its impact, it is a peek into a potential pathway for the percutaneous treatment of renal and potentially other atherosclerotic stenotic lesions. Since the primary aim of the treatment is preservation of end-organ function and stabilization of end-organ physiology, whether it be in the brain, heart, kidney or calf muscle, any clinically significant liberation of embolic debris during the therapeutic procedure could be potentially detrimental to the intended benefit. This is so clear cut in the brain, which is an eloquent organ, that the concept of embolic protection during carotid stenting is no longer in question. Indeed, it is accepted as standard in saphenous vein graft intervention, and has become increasingly the norm in peripheral vascular atherectomy in patients with only a single runoff vessel.
For renal intervention, the question is not yet answered. The studies of Henry and Amor with balloon and Holden with filter protection, alluded to above, suggest an advantage. The RESIST trial interestingly suggested that while filter protection or treatment with abciximab alone decreased the amount of loss of glomerular filtration, only the combination of both anti-platelet therapy and filtration resulted in an increase in GFR post procedure.
For those of us who do renal intervention, it is clear that in the majority of cases, benefit may be achieved with the relatively straightforward approach of protecting the renal bed against aortic embolization using the no-touch technique originally described by Feldman and a direct stenting approach without pre- or post-dilatation. This method allows a rapid completion of the procedure, minimizes contrast use and lesion trauma. As an example, if one looks at the number of cardiac cycles with intrarenal microembolic signals registered by Kawarada's group after predilatation and postdilatation, they sum to 11.6 whereas after stent balloon deflation they are seen in 11.1, so one might infer that the embolic load could be halved by a strategy of stenting alone rather than predilatation and postdilatation in addition to stenting. These simple approaches can be used to great effect in patients without angiographic evidence of thrombus, and avoid the excess cost without proven benefit of filter or balloon protection devices.
Thus, although it is clear that during interventions in the renal bed, as in the coronary, peripheral and carotid circulations, embolic phenomena are often encountered, large randomized studies are lacking to show strong evidence in favor of routine use of expensive adjunct protection strategies. The study in this issue of the Journal suggests that IVUS with virtual histology may prove a useful method for interrogating lesions prior to the decision to protect with a filter by identifying those lesions most likely to embolize and decrease renal blood flow after the intervention. That said, the very act of passing the lesion with the IVUS catheter may induce embolization and is in itself costly. Alternative methods such as plaque characterization using gray scale median analysis of surface ultrasound, and MRI may be helpful in identifying those lesions most prone to cause embolization and further renal dysfunction.
Abstract and Introduction
Introduction
Surgical correction of renal arterial disease has been reported in the United States since 1975 and was primarily used in patients suffering severe or refractory hypertension or incipient renal failure with demonstrable critical renal artery stenoses. This method has largely been discarded in favor of percutaneous intervention, initially with balloon angioplasty alone and more recently with stent implantation. Despite the enthusiasm for correction of the anatomic lesion thought to be responsible for hypertension and loss of renal parenchymal function, many patients do not reap the intended benefit, either due to a misunderstanding of the physiologic significance of the treated lesion, progression of intrinsic renal disease, continued aortic atheroembolism or complications of the procedure itself. Because of these findings, renal artery revascularization for atherosclerotic renal arterial stenosis remains controversial, with a recent publication in the New England Journal of the ASTRAL study concluding that revascularization offers "no worthwhile clinical benefit" over medical therapy.
Of the feared sequelae of renal revascularization, embolization has been most often cited as a reason for continued renal tissue loss and decline of function. Small studies have suggested that some degree of embolization is virtually ubiquitous during renal stent implantations, and these findings are corroborated by the presence of emboli in debris aspirated from balloon-protected renal stents as well as trapped in filter protected renal stents. Suppose we were able to predict on the basis of pre-stent imaging which lesions were most likely to result in liberation of debris and worsening of renal function? A tantalizing report in this issue of the Journal by Prasad et al proposes to do just that. Using intravascular ultrasound "virtual histology," a method utilizing radiofrequency backscatter to identify the contents of arterial plaque, the authors relate the presence of necrotic core with a worsening of renal arterial flow after renal artery stenting as judged by renal frame count – the number of frames required for radiocontrast to reach the end arteries of the kidney. An increase in frame count means that contrast takes longer to fill the renal bed and is a harbinger of a poor result after renal intervention.
Using a clinically available intravascular ultrasound (IVUS) system with a phased array catheter and a virtual histology (VH) software package, the authors measured the plaque constituents, imaging throughout a vessel length beginning 5mm distal to the stented segment by continuous motorized pullback to the ostium of the renal artery. VH representations of fibrous tissue, fibrofatty tissue, necrotic core and dense calcifications were volumetrically quantified throughout the assessed vessel length and at the point of minimal luminal diameter within the lesion. The clinical features of the patients were tabulated, and interestingly, similarly to other studies of interventional patients, only 59% of those studied had been treated with HMG-CoA reductase inhibitors at the time of intervention. Patients were accessed using the no-touch technique, treated with routine predilatation, which is not agreed standard practice, and then stented. IVUS was performed prior to predilatation in 17 of the 18 patients. One lesion was so tight that it was judged to require predilatation prior to passing the stenosis with the IVUS catheter. One lesion was treated without stenting – we are not told why. Two lesions were not interrogated using continuous pullback, and only the minimal luminal diameter frame was analyzed in these two patients. No protection was used in this cohort.
Using this technique, the renal artery stenoses were reduced from a mean of 66% to 8.6%, and renal frame count for the overall group was reduced from 34.5 to 30.5 frames (mean reduction 2.5 ± 13 frames). 10 out of the 17 patients showed improvement in flow with a mean reduction in renal frame count of 11.9 ± 7.6 frames. The other 7 patients showed worsening of flow with a mean increase of 7.1 ± 8.0 frames.
There were two important findings of this study. First, analysis of the IVUS determined virtual histology showed that the stented segments were primarily composed of fibrous tissue (62.1%), with less necrotic core (15.6%), dense calcification (13.2%) and fibrofatty tissue (8.7%). Second, and most important, there was a clear relationship seen between an increase in the amount of necrotic core and an increase in renal frame count after stenting. In other words, the more necrotic core seen in the stented segment, the greater the increase in renal frame count, and thus the greater the loss of renal flow.
This was a small group of patients, only 18, and although a reduction in systolic blood pressure was seen over the whole group at 10.6 weeks in all patients (141 ± 22 mmHg baseline vs. 124 ± 8.3 mmHg at follow-up) no statements could be made about the relationship of plaque composition or renal frame count and the clinical outcomes of blood pressure control or change in GFR.
Although this was a small study, and therefore limited in the scope of its impact, it is a peek into a potential pathway for the percutaneous treatment of renal and potentially other atherosclerotic stenotic lesions. Since the primary aim of the treatment is preservation of end-organ function and stabilization of end-organ physiology, whether it be in the brain, heart, kidney or calf muscle, any clinically significant liberation of embolic debris during the therapeutic procedure could be potentially detrimental to the intended benefit. This is so clear cut in the brain, which is an eloquent organ, that the concept of embolic protection during carotid stenting is no longer in question. Indeed, it is accepted as standard in saphenous vein graft intervention, and has become increasingly the norm in peripheral vascular atherectomy in patients with only a single runoff vessel.
For renal intervention, the question is not yet answered. The studies of Henry and Amor with balloon and Holden with filter protection, alluded to above, suggest an advantage. The RESIST trial interestingly suggested that while filter protection or treatment with abciximab alone decreased the amount of loss of glomerular filtration, only the combination of both anti-platelet therapy and filtration resulted in an increase in GFR post procedure.
For those of us who do renal intervention, it is clear that in the majority of cases, benefit may be achieved with the relatively straightforward approach of protecting the renal bed against aortic embolization using the no-touch technique originally described by Feldman and a direct stenting approach without pre- or post-dilatation. This method allows a rapid completion of the procedure, minimizes contrast use and lesion trauma. As an example, if one looks at the number of cardiac cycles with intrarenal microembolic signals registered by Kawarada's group after predilatation and postdilatation, they sum to 11.6 whereas after stent balloon deflation they are seen in 11.1, so one might infer that the embolic load could be halved by a strategy of stenting alone rather than predilatation and postdilatation in addition to stenting. These simple approaches can be used to great effect in patients without angiographic evidence of thrombus, and avoid the excess cost without proven benefit of filter or balloon protection devices.
Thus, although it is clear that during interventions in the renal bed, as in the coronary, peripheral and carotid circulations, embolic phenomena are often encountered, large randomized studies are lacking to show strong evidence in favor of routine use of expensive adjunct protection strategies. The study in this issue of the Journal suggests that IVUS with virtual histology may prove a useful method for interrogating lesions prior to the decision to protect with a filter by identifying those lesions most likely to embolize and decrease renal blood flow after the intervention. That said, the very act of passing the lesion with the IVUS catheter may induce embolization and is in itself costly. Alternative methods such as plaque characterization using gray scale median analysis of surface ultrasound, and MRI may be helpful in identifying those lesions most prone to cause embolization and further renal dysfunction.
Source...