Renal artery stenosis: CT angiography or gadolinium MR angiography are best at diagnosing it

Clinical bottom line (level 2a)

  1. Roughly a third of patients with suspected renovascular hypertension have renal artery stenosis.
  2. CT angiography or gadolinium-enhanced MR angiography can help diagnose it or rule it out.
  3. Abnormal ultrasonography or captopril renal scintigraphy makes renal artery stenosis more likely, and normal tests make it less likely. However further testing is required.
  4. The captopril test is not very helpful at diagnosing or ruling it out.
Vasbinder et al: Ann Intern Med 2001; 135 : 401-411
Expires May 2004

The study

Systematic review of all diagnostic studies of
  • Patients: suspected renovascular hypertension
  • Intervention: CT angiography, gadolinium-enhanced 3-D MR angiography, ultrasonography, captopril renal scintigraphy, captopril test compared with intra-arterial X-ray angiography
  • Outcome: renal artery stenosis (50%-60% stenosis)

Articles found in English, French, German using Medline, Embase, Cochrane, to August 2000 (search terms: detailed in text ) and hand-searching bibliographies of retrieved articles

Selection criteria: by 2 independent reviewers - disagreements were resolved by consensus
Appraisal criteria: by 2 independent reviewers: using an independent reference standard in all patients; assessed for blinding of reference standard
Articles excluded if:
  • criteria and cut-off values for a positive result not given
  • true positive, true negative, false positive, false negative rates not given
  • not all patients underwent the reference standard
55 studies found
  • 5 using CT angiography
  • 16 using MR angiography
  • 14 using captopril renal scintigraphy
  • 4 using the captopril test

Study results were found to be heterogeneous due in part to the different criteria for a positive test.

The evidence


diagnostic test renal artery stenosis no renal artery stenosis LR+
(95% CI)
post-test probability LR-
(95% CI)
post-test probability
gadolinium-enhanced MR angiography 16
(10 to 26)
90% 0.019
(0.0062 to 0.059)
1%
CT angiography 26
(16 to 41)
93% 0.037
(0.020 to 0.068)
2%
ultrasonography 8.0
(6.9 to 9.3)
81% 0.18
(0.16 to 0.21)
9%
captopril renal scintigraphy 4.5
(3.9 to 5.2)
71% 0.19
(0.15 to 0.23)
9%
captopril test 2.8
(1.9 to 3.9)
60% 0.69
(0.58 to 0.81)
27%
total

  • CT angiography and gadolinium-enhanced MR angiography were found to be equally accurate, and significantly better than the other tests on ROC analysis. Other MR angiographic techniques were less accurate than gadolinium-enhancement.
  • The captopril test was found to be significantly worse than any other test on ROC analysis.
  • Prevalence ranged from 25% to 38%. 35% used to calculate the post-test probabilities above.

Citation

  1. Vasbinder GB, Nelemans PJ, Kessels AG, et al: diagnostic tests for renal artery stenosis in patients suspected of having renovascular hypertension: a meta-analysis. Ann Intern Med 2001; 135 : 401-411
Search Terms: from ACP Journal Club
Contributor: Chris Ball, May 2002
Reviewer:

Clinical Question.
Patient suspected renovascular hypertension
Intervention or Exposure CT angiography, MR angiography, ultrasound, captopril renal scintigraphy, captopril test
Comparison contrast angiography
Outcome renal artery stenosis