| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Department of Radiology (B1D 502), University of Michigan Medical Center, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0030. From the 2001 RSNA scientific assembly. Received September 24, 2001; revision requested December 3; final revision received June 7, 2002; accepted June 28. Address correspondence to R.O.B. (e-mail: ronbude@umich.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: In 68 adults (24 men, mean age, 67 years ± 10; 44 women, mean age, 67 years ± 12), angiograms that fulfilled the following criteria were studied: (a) technically adequate renal angiograms obtained to evaluate suspected renovascular hypertension and (b) angiographically documented hemodynamically significant stenosis of any renal artery. The percentage of kidneys and the percentage of patients with hemodynamically significant isolated stenoses of accessory renal arteries were calculated.
RESULTS: Eighty-seven kidneys in 68 patients had hemodynamically significant renal artery stenoses. Fifteen kidneys had 16 accessory renal arteries. Four accessory arteries in three patients had hemodynamically significant stenoses. Only one of 68 patients (1.5%) had an accessory artery stenosis unaccompanied by a main renal artery stenosis in either kidney; this patient had bilateral hemodynamically significant accessory artery stenoses. Two patients had coexistent hemodynamically significant stenoses of accessory and main renal arteries.
CONCLUSION: The prevalence of a hemodynamically significant stenosis isolated to an accessory renal artery was 1.5% in our study. Thus, failure to detect accessory renal arteries should not unduly affect the utility of a noninvasive test for detecting renovascular hypertension.
© RSNA, 2003
Index terms: Hypertension, renovascular, 961.723 • Kidney, blood supply, 961.721 • Renal angiography, 961.122 • Renal arteries, 961.1312 • Renal arteries, stenosis or obstruction, 961.721
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Accessory arteries occur in at least 15% of kidneys (3–6). An isolated accessory renal artery stenosis, with patent main renal arteries, can produce renovascular hypertension (5,7). In a recent study, only 21% of accessory renal arteries were depicted with unenhanced ultrasonography (US) (4); investigators in earlier studies were even less successful (5,6). When these factors are considered together, it is often assumed that US is substantially less effective as a noninvasive means for detecting renovascular hypertension because of its inability to reliably depict accessory renal arteries.
It seemed to us, however, that this line of reasoning might be incorrect. In the noninvasive evaluation of renovascular hypertension, it is not important how often accessory arteries are present, nor is it of prime importance how often they have a hemodynamically significant stenosis. With the assumption that the noninvasive test in question detects hemodynamically significant main renal artery stenoses to an acceptable degree, it is only important how often a hemodynamically significant stenosis involves an accessory renal artery in isolation (ie, both main renal arteries are patent). This is because virtually all patients who have kidneys with renal artery stenoses suspected of being hemodynamically significant proceed to angiography, not only to confirm the presence of the stenosis, but also to enable the use of angioplasty or stent placement to treat the stenosis when it is to be treated nonsurgically or to completely delineate the arterial anatomy when surgery is to be performed. Thus, even if the results of noninvasive examination do not detect an accessory renal artery when there is a hemodynamically significant stenosis of the main renal artery, angiography should still depict the accessory artery later in the work-up. If the accessory artery has a hemodynamically significant stenosis, it can still be treated angiographically or surgically at that time, with no ill effect or delay in treatment to the patient.
This study was performed to determine how often stenoses of hemodynamic significance involve accessory renal arteries in isolation (ie, with patent main renal arteries).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
We desired our study population to consist only of patients with proved hemodynamically significant renal artery stenoses who underwent angiographic work-up for suspected renovascular hypertension. To achieve this, a computer search of the entire duration of our radiology database (5 years 10 months) was used to identify all renal angiograms (682 patients). The angiographic reports and, when necessary, the medical records of all 682 patients were reviewed by one of the authors (R.O.B.) to identify only those adults undergoing work-up for renovascular hypertension who had one or more stenoses in any renal artery that were likely to be hemodynamically significant. Those patients who had undergone prior abdominal aortic or renal artery surgery or angioplasty were excluded. The angiograms of all individuals (n = 130) satisfying these initial criteria were reviewed; only those patients whose angiograms satisfied the following criteria formed the final study population: (a) hemodynamically significant stenoses in the main renal artery, its branches, or an accessory renal artery; (b) angiograms that were technically adequate to identify and evaluate accessory renal arteries; and (c) no prior aortic or renal artery surgery, stent procedures, or angioplasty. A detailed description of the method follows.
Review of reports of initial computer-generated population of 682 patients with renal angiograms.—The reports of all of these angiographic studies and, when necessary, the medical records were reviewed, and the following inclusion and exclusion criteria were applied:
1. Only adults (
18 years old) who had undergone angiography for suspected renovascular hypertension were included. Any patient with a renal artery stenosis incidentally noted in the work-up for any other condition, such as renal tumor or aortic aneurysm, was excluded from the study because we wished our study population to mirror the population undergoing evaluation for renovascular hypertension only.
2. Any patient who had undergone prior abdominal aortic or renal artery surgery or a prior renal artery angioplasty or stent procedure was excluded.
3. Only those patients with angiographic reports describing stenoses that were likely to be hemodynamically significant in any portion of the renal arterial system, including in an accessory artery only, were included. The following initial criteria for hemodynamic significance were applied (hemodynamic significance was reassessed at review of the angiograms; to be discussed subsequently): (a) transstenotic systolic pressure gradient of 10 mm Hg or more, (b) stenosis of 50% or more of luminal diameter, (c) any stenosis described as greater than "mild" when only verbal descriptors were used, or (d) the presence of fibromuscular dysplasia.
4. Patients with total renal artery occlusions had data from the occluded kidney excluded from analysis. In these cases, a noninvasive test is usually inconclusive because a renal artery is not identified, making it difficult to determine whether the artery is occluded or is not identified because of sampling error or technical difficulty; further work-up of the patient would likely be required, and any coincidental stenosis of an accessory renal artery could be evaluated at that time. Application of these criteria resulted in the identification of a smaller group of 130 patients, whose angiograms were then reviewed for adequacy of evaluation for accessory arteries and also to better evaluate the hemodynamic significance of the renal artery stenoses.
Review of angiograms of group of 130 patients to arrive at final study population.—The angiograms of all 130 patients were reviewed by an experienced angiographer (A.R.F.) blinded to all clinical information. The following criteria were applied:
1. All patients with evidence of abdominal aortic or renal artery surgery were excluded because the surgery likely altered the anatomic structures. This step was necessary because not all surgical procedures were mentioned in the reports.
2. All patients with renal artery stents had the data from that kidney excluded because of the prior intervention to that kidney. These patients were included in the study only if they had a hemodynamically significant stenosis of the contralateral untreated kidney.
3. All patients with technically inadequate angiographic examinations to evaluate for the presence of accessory renal arteries were excluded. Technical adequacy was determined in the following way. If the angiographic report clearly stated the number of arteries to each kidney (as determined by a reviewer other than the angiographer, because the angiographer was blinded to the reports), the patient was included in the study. If the angiographic report did not state the number of arteries to each kidney, the angiograms were reviewed by an angiographer (A.R.F.) to determine whether they were technically adequate to evaluate for accessory renal arteries; if not, the patient was excluded. Three patients were included in the final study population who had technically inadequate angiograms of one kidney but had a hemodynamically significant stenosis of the renal artery of the contralateral adequately evaluated kidney. All other patients included in the study had technically adequate angiograms of both kidneys.
4. Only those patients with hemodynamically significant renal artery stenoses were included. Initially, this criterion might seem redundant because the angiographic reports had already been checked for hemodynamically significant renal artery stenoses; however, the reports were not always definitive for the presence of hemodynamically significant stenoses. This is because when transstenotic systolic pressure gradients were not measured, the stenosis was often graded either qualitatively (mild, moderate, or severe) or with an estimation, instead of an actual measurement, of the degree of luminal narrowing (percentage of diameter). The following criteria were strictly applied to all of these angiographic studies: (a) A transstenotic systolic pressure gradient of 10 mm Hg or more was considered the best proof of a hemodynamically significant stenosis and took precedence over any other means of quantifying stenosis. (b) When a transstenotic systolic pressure gradient was not measured, the narrowing of the luminal diameter was measured with hand calipers; any stenosis of 50% or more was included.
Except for patients with fibromuscular dysplasia, these two criteria sufficed. Individuals with fibromuscular dysplasia can have hemodynamically significant stenoses even if the visual degree of luminal narrowing of any of the often many areas of stenosis does not exceed 50%. This is because focal webs of hemodynamic significance can be difficult to image. In addition, a series of individually insignificant stenoses, all less than 50% in luminal diameter reduction, can exert a cumulative effect that is hemodynamically significant. Pressure gradients are not usually measured in these individuals because of the risk of catheter-induced arterial dissection inherent in this disorder. The decision to treat these patients is often empirically made and is based not only on the clinical history but also on the overall "gestalt" of the degree of arterial involvement.
The following criteria determined which patients with fibromuscular dysplasia to include of those in whom transstenotic systolic pressure gradients were not determined and in whom the luminal narrowing of the greatest degree of stenosis was less than 50%. If a stent procedure or angioplasty of the involved artery had been performed, the patient was included in the study; if the artery had not been treated and if the hypertension was treated only medically, the patient was excluded. Although this introduced a subjective element into our study, it seemed best to include these patients because fibromuscular dysplasia is encountered fairly frequently, and many angiographers will treat these patients by using criteria similar to ours.
Of the 13 patients (all women) with fibromuscular dysplasia in the initial group of 130 patients, six had not undergone a stent procedure or angioplasty and did not have main renal artery luminal narrowing of 50% or more or have transstenotic systolic pressure gradients determined. These six were excluded from the study. Of the remaining seven patients with fibromuscular dysplasia, three had stenoses with either (a) a transstenotic systolic pressure gradient that was measured and was 10 mm Hg or more or (b) a luminal diameter narrowing of 50% or more; the remaining four patients neither had transstenotic systolic pressure gradients determined nor had any luminal diameter narrowings of 50% or more but had undergone angiographic therapy (stent procedure or angioplasty) and were therefore included in the final study population.
After application of all of these criteria, 62 individuals were excluded, giving a final study population of 68 patients with a hemodynamically significant stenosis in at least one renal artery. Of the final 68 patients, 24 were men, and 44 were women. The mean age of the men was 67 years ± 10 (range, 39–83 years); the mean age of the women was 67 years ± 12 (range, 34–84 years).
Angiographic Studies
Cut-film or digital subtraction angiography had been performed in all patients. Angiography had been performed in 56 patients with iodinated contrast agents, in eight patients with gadolinium-based contrast agents, and in four patients with carbon dioxide. Aortograms had been obtained in the anteroposterior or shallow left anterior oblique projection by using 4- or 5-F flush catheters. Selective renal angiograms had been obtained by using 4- or 5-F catheters.
Data Analysis
The percentage of kidneys and the percentage of patients with hemodynamically significant isolated accessory renal artery stenoses were calculated. The large-sample 95% CI for the proportion of patients with hemodynamically significant isolated accessory renal artery stenoses, assuming a binomial distribution, was calculated.
Noninvasive Diagnostic Tests before Angiography
Often a noninvasive renal examination precedes angiography in the work-up of renovascular hypertension. With the exclusion of eight patients in our final study population for whom the initial diagnostic procedure, if any, was unknown (they were referred from outside hospitals, and our records were incomplete), 39 patients underwent such initial noninvasive diagnostic procedures: gadolinium-enhanced magnetic resonance (MR) angiography (n = 35), captopril renography (n = 3), and US (n = 1). Thus, 65% (39 of 60 patients) of the individual patients in our final study population for whom these data are known underwent a noninvasive test preceding angiography; of these, MR angiography was performed in 90% (35 of 39 patients). The remaining 21 patients underwent aortography and/or renal angiography as the sole diagnostic test in their work-ups for renovascular hypertension; 16 patients were evaluated angiographically for renovascular hypertension and another condition concurrently (coronary artery disease or peripheral vascular disease), while five patients underwent primary renal angiography alone. Thus, 35% (21 of 60 patients) of our final study population for whom we have complete information underwent angiography as the sole diagnostic test for the work-up of renovascular hypertension.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Of 77 kidneys with atherosclerotic stenoses in 61 patients, the predominant stenosis involved only the main renal artery in 70 of the 77 kidneys (91%). In another kidney, the main renal artery (4-mm diameter) had an early bifurcation 5 mm after its origin and divided into coequal branches 2.8 mm in diameter; the more caudal of the two had a 71% atherosclerotic stenosis at its origin (contralateral kidney without hemodynamically significant stenosis). In four kidneys of another three patients, atherosclerotic stenoses involved accessory arteries (discussed subsequently). Finally, two patients had unilateral hemodynamically significant stenoses of main renal arteries and also ipsilateral accessory arteries that could not be seen well enough to determine if they had hemodynamically significant stenoses (see Discussion). There were no patients in whom the main renal artery was free of a hemodynamically significant stenosis and in whom an accessory artery could not be evaluated adequately for a hemodynamically significant stenosis.
Hemodynamically significant stenoses were found in 10 kidneys involved by fibromuscular dysplasia. Only two of these kidneys (in one patient) had accessory arteries; both of these accessory arteries were without any stenosis.
Unilateral kidneys in 11 of the 68 patients were excluded from analysis (these were contralateral to kidneys with hemodynamically significant renal artery stenoses). Four of these 11 kidneys had preexistent renal artery stents; in one kidney, prior angioplasty had been performed in a renal artery; in three kidneys, the renal artery was completely occluded; and in three kidneys, the angiogram was technically inadequate to evaluate for accessory arteries.
Only one of the 68 patients (1.5%) had a hemodynamically significant stenosis of an accessory renal artery while both main renal arteries were without a hemodynamically significant stenosis. This patient actually had bilateral accessory arteries, each with a hemodynamically significant stenosis (54% and 59% luminal diameter narrowing) that was subsequently treated with angioplasty. The large-sample 95% CI for the proportion of patients with hemodynamically significant isolated accessory renal artery stenoses was determined (95% CI: 0.00, 0.043). Two of the 68 patients had a hemodynamically significant stenosis of an accessory renal artery but also had a hemodynamically significant stenosis of a main renal artery; one had a hemodynamically significant main renal artery stenosis on the same side as the accessory artery stenosis, and the other had a hemodynamically significant main renal artery stenosis contralateral to the accessory artery stenosis. The distribution and characterization of the renal artery stenoses in the 12 patients with hemodynamically significant stenoses and accessory arteries are fully delineated in the Table.
Stenosis and Pressure Gradients
The severity of stenosis was determined with transstenotic systolic pressure gradient measurements in 42 patients (53 kidneys) and by using the percentage of luminal diameter narrowing in 22 patients (28 kidneys). Four patients (six kidneys) were included who had fibromuscular dysplasia that was subsequently treated with angioplasty or a stent procedure but was not quantified with transstenotic systolic pressure gradient measurements before therapy and did not demonstrate a luminal diameter narrowing of 50% or more of the involved artery.
Transstenotic systolic pressure gradients had been measured in 16 patients in the initial study population of 130 patients who had unilateral main renal artery stenoses in the 50%–59% range. In 13 of the 16 patients (81%), the transstenotic systolic pressure gradient was 10 mm Hg or more (mean, 69 mm Hg ± 44; range, 15–150 mm Hg); in the remaining three patients, the transstenotic systolic pressure gradient was less than 10 mm Hg. These three patients were excluded from the final study population because pressure gradients had the highest priority as a criterion for inclusion or exclusion from the study, and the gradients of these patients were below threshold.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
One of the 87 kidneys had neither a hemodynamically significant stenosis of the main artery nor such stenosis of an accessory artery; the kidney also did not have any accessory arteries. This was the kidney with stenosis of 71% luminal narrowing at the origin of one branch of an early bifurcation of the main renal artery. Because the stenosis originated only 5 mm from the origin of the main renal artery and was present in one of two coequal branches of the main renal artery, this patient was included in the group of main renal artery stenoses because, for detection purposes, this is essentially equivalent to a main renal artery stenosis.
Finally, in two kidneys (two patients), there was a hemodynamically significant stenosis of the main renal artery, and each kidney had accessory renal arteries that could be identified, but these accessory arteries could not be seen well enough to determine whether a hemodynamically significant stenosis was present or not (Table). Therefore, we cannot exclude hemodynamically significant stenoses of these accessory arteries; however, this is unimportant for noninvasive evaluation, as already discussed.
|
Our results show that an isolated hemodynamically significant stenosis of an accessory renal artery is an infrequent finding, occurring only once in the 68 patients in our study (1.5%). This prevalence is so low that we believe it can be ignored in the noninvasive evaluation of renovascular hypertension; it suggests that the utility of a noninvasive test for the detection of renovascular hypertension can rest solely on the ability of the test to detect hemodynamically significant stenoses of the main renal artery.
The results of this study are specifically important for US. They show that for renovascular hypertension, the utility of US should currently rest solely on its ability to help evaluate the main renal artery. In addition, our results likely err on the conservative side, for two reasons. First, unenhanced US can be used to depict some accessory renal arteries; Halpern et al (4) depicted 21% of them. As US contrast agents become available, this detection rate should improve. Thus, US should even be able to be used to depict some hemodynamically significant accessory renal artery stenoses. Second, not all patients who have accessory renal arteries with a hemodynamically significant stenosis undergo a procedure to correct the stenosis. In our institution, small accessory arteries with a hemodynamically significant stenosis are sometimes not treated with angioplasty, a stent procedure, or surgery and are treated pharmacologically instead. If a noninvasive modality fails to detect one of these stenoses, it is likely of little consequence to the patient.
Viewed in a slightly different way, our results can be considered as follows. If renovascular hypertensive patients constitute approximately 5% of the hypertensive population, and if approximately 1.5% of renovascular hypertensive patients have an isolated hemodynamically significant accessory renal artery stenosis, and if all accessory arteries go undetected, noninvasive evaluation for renovascular hypertension will not allow detection of the responsible accessory artery stenosis in only approximately 0.08% of the population (1.5% x 5%). Therefore, it seems to us that it is unnecessary to expend much effort to evaluate the accessory arteries for a hemodynamically significant stenosis. This may not only provide impetus for the acceptance of US as an adequate noninvasive modality but also help to reduce examination time if extreme diligence does not need to be devoted to the search for accessory arteries.
Limitations of our study include the following. First, some subjects undergoing evaluation for renovascular hypertension first undergo a noninvasive test (MR angiography, captopril renography, computed tomographic angiography, or US). Noninvasive tests such as US and renography do not depict isolated stenosis of an accessory renal artery well, and when these tests are used to evaluate renovascular hypertension, such a false-negative result for a hemodynamically significant renal artery stenosis might cause the work-up to terminate without including angiography, when angiography might have depicted an isolated accessory renal artery stenosis. This might have biased our final study population against inclusion of patients with stenoses of accessory arteries only and caused an underreporting of the prevalence of such stenoses; however, we do not believe this to be a substantial limitation to our study for the following reasons: Only 65% of our final study population for whom there was complete information underwent an initial noninvasive test in their evaluations for renovascular hypertension; the remaining 35% underwent angiography as the sole diagnostic test in their work-ups for renovascular hypertension.
These same percentages should apply to those patients in our study who were initially evaluated for renovascular hypertension, who were concluded not to have a hemodynamically significant renal artery stenosis, and who were excluded, leaving our final study population of 68 patients. Therefore, approximately 35% of individuals who were concluded not to have hemodynamically significant renal artery stenoses probably also underwent angiography as their primary diagnostic test and thus were not subjected to this possible limitation.
Of the remaining 65% of individuals who underwent a primary noninvasive test and were concluded not to have a hemodynamically significant renal artery stenosis on the basis of results of that test alone, it is reasonable to assume that for approximately 90% of them, this test was gadolinium-enhanced MR angiography (because 90% of our final study population underwent gadolinium-enhanced MR angiography first). Numerous studies have shown that gadolinium-enhanced MR angiography can be used to depict at least 90% of accessory renal arteries (12–16), and early evidence suggests that gadolinium-enhanced MR angiography might be useful for depicting hemodynamically significant accessory renal artery stenoses; six of six accessory arteries with hemodynamically significant stenoses were depicted in two recent studies (14,15). Therefore, even though we cannot absolutely prove that our study was not biased against the detection of isolated hemodynamically significant accessory renal artery stenoses, we believe that the aforementioned data strongly suggest that any such bias was minimal and not substantial enough to affect our conclusions in a meaningful way.
A second limitation of our study is that the main inclusion criterion was a renal artery stenosis that was hemodynamically significant according to published criteria. Clinical outcomes were not considered, and it is well known that elimination of a stenosis considered to be hemodynamically significant does not necessarily result in a lessening of hypertension in a patient. However, this is a limitation not only of our study, but of the evaluation for renovascular hypertension in general, because it is impossible to reliably predict which patients will benefit from correction of their stenoses. Third, our population of patients may not exactly match those of other institutions. However, one of the entrance requirements of our study was that the patients were undergoing evaluation for renovascular hypertension. We do not know of any factors that made our population substantially different from those of other institutions.
We conclude that the success of a noninvasive modality for detecting renovascular hypertension is essentially solely dependent on the ability of that modality to evaluate the main renal artery for hemodynamically significant stenoses. Any ability of that modality to detect hemodynamically significant stenoses of accessory arteries provides additional information but should not substantially improve the rate of detection of renovascular hypertension with that modality. This is especially important for US, given its current inability to reliably depict accessory renal arteries. The utility of US in the evaluation of renovascular hypertension should be based on its ability to evaluate the main renal artery, which may improve to a universally acceptable level with the use of US contrast agents.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Soulez, M. Pasowicz, G. Benea, L. Grazioli, J. P. Niedmann, M. Konopka, P. C. Douek, G. Morana, F. K. W. Schaefer, A. Vanzulli, et al. Renal Artery Stenosis Evaluation: Diagnostic Performance of Gadobenate Dimeglumine-enhanced MR Angiography--Comparison with DSA Radiology, April 1, 2008; 247(1): 273 - 285. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Maki, G. J. Wilson, W. B. Eubank, D. J. Glickerman, J. A. Millan, and R. M. Hoogeveen Navigator-Gated MR Angiography of the Renal Arteries: A Potential Screening Tool for Renal Artery Stenosis Am. J. Roentgenol., June 1, 2007; 188(6): W540 - W546. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Willoteaux, M. Faivre-Pierret, O. Moranne, C. Lions, J. Bruzzi, M. Finot, V. Gaxotte, C. Mounier-Vehier, and J.-P. Beregi Fibromuscular Dysplasia of the Main Renal Arteries: Comparison of Contrast-enhanced MR Angiography with Digital Subtraction Angiography Radiology, December 1, 2006; 241(3): 922 - 929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gupta and R. Tello Accessory Renal Arteries Are Not Related to Hypertension Risk: A Review of MR Angiography Data Am. J. Roentgenol., June 1, 2004; 182(6): 1521 - 1524. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
