Captopril test: time over?

Transcription

1 ORIGINAL ARTICLE Journal of Human Hypertension (1999) 13, Stockton Press. All rights reserved /99 $ T Lenz 1, T Kia 1, G Rupprecht 1, K-L Schulte 1,2 and H Geiger 1 1 Division of Nephrology, Medical Clinic IV, Johann Wolfgang Goethe University, Frankfurt am Main, Germany Objective: The role of non-invasive tests for the detection of renovascular hypertension is still a matter of controversy. The captopril test is widely used; its clinical usefulness, however, remains questionable. The aim of the current study was therefore to report our own experience and to review the published data on the diagnostic significance of the test. Patients and methods: Data from 485 hypertensive patients who underwent a captopril test in consecutive order at our institution were analysed retrospectively. After a 30-min resting period in the supine position 50 mg of captopril was given orally. Blood was collected before and 90 min after dosage for the determination of plasma renin concentration (normal range ng/ml/h). An increase by 100% or more of the baseline value was considered a positive response. Blood pressure was recorded at baseline and at 90 min. Results: A positive response was present in 62 patients; further diagnostic work-up revealed significant renal artery stenosis in 11 of these patients. In the 423 patients with a negative response renal artery stenosis was found in three cases. With some limitations of retrospective analyses in mind, sensitivity and specificity of the test were calculated as 79% and 89%, respectively. No severe complication occurred during the test. Conclusion: Our data on the diagnostic indices and the safety of the captopril test are in good agreement with most published series. Altogether, available data suggest that the captopril test has a limited diagnostic accuracy as a screening test for the detection of renovascular hypertension. New radiologic non-invasive techniques with greater diagnostic value are therefore likely to challenge the clinical role of the test in the future. Keywords: captopril test; renovascular hypertension; magnetic resonance imaging; computerised tomography angiography Introduction Renovascular hypertension is the most frequent form of secondary hypertension, although its precise prevalence is not known. Several competing noninvasive diagnostic procedures have been proposed but there is no generally accepted consensus on how to proceed in patients with suspected renovascular disease. Case and Laragh 1 were the first to show that the reactive rise of plasma renin after an angiotensin converting enzyme (ACE) inhibitor is greater in patients with renovascular hypertension than with essential hypertension. This observation led to the introduction of the so-called captopril test for the diagnostic work-up of patients with suspected renovascular hypertension. At first quite enthusiastic results with high sensitivity (up to 100%) and specificity were reported 2 ; subsequent studies, however, found lower diagnostic accuracy. It has been suggested that the differences in the reported accuracy may be attributable to several factors such as the renin assay method used, varying criteria for defining renovascular hypertension, sodium restriction and position of the patient during the test (for review see reference 3). A decade ago Gaul and Correspondence: Dr Tomas Lenz, Division of Nephrology, Medical Clinic IV, Johann Wolfgang Goethe University, Theodor Stern Kai 7, Frankfurt am Main, Germany 2 Current address: Medical Clinic, Königin Elisabeth Hertberge Hospital, Berlin Received 26 October 1998; accepted 22 January 1999 coworkers 4 searched the published studies on the captopril test and selected 16 of them for further analysis according to certain quality criteria. At that time these authors concluded that the clinical usefulness of the test cannot be assessed on the basis of the existing data since clear cut-off points were not defined and the test results were not always confirmed by arteriography. Therefore, in the current paper the literature of the ensuing 10 years including our own experience is presented in order to answer the question whether we have gained important new insights on the usefulness of the captopril test in the last decade. Patients and methods Patients Medical records from 485 patients (242 female, 243 male; mean age 44.7 ± 13.3 (range 10 81) years) with suspected renovascular hypertension who in consecutive order underwent a captopril test at our institution from 1991 through 1993 were reviewed. According to the available blood pressure measurements in the patient charts prior to the testing (total n = 457) 43% of the patients were borderline hypertensive (sitting diastolic blood pressure mm Hg), 50% were moderately hypertensive (sitting diastolic blood pressure mm Hg) and 7% were severely hypertensive (sitting diastolic blood pressure 115 mm Hg).

2 432 Captopril testing procedure Diuretics and ACE inhibitors were discontinued at least 1 week before testing; if possible no antihypertensive medication was given 12 h prior to testing. Patients were advised to maintain a normal salt intake. After a 30-min resting period in the supine position 50 mg of captopril was given orally with some tap water. Blood (10 ml) was collected and placed on ice before and 90 min after dosing for the determination of plasma renin concentration. An increase by 100% or more of the baseline value was considered a positive response. Blood pressure was recorded at baseline and at 90 min. Plasma renin concentration was determined with a variation to the method by Boucher et al, 5 using nephrectomized sheep renin substrate (normal range ng/ml/h). Statistical analysis The results are given as mean ± s.e.m. Wilcoxon Mann Whitney statistics were used for comparisons between groups. P-values 0.05 were considered significant. Results Renin response/diagnostic indices In the whole group plasma renin concentration amounted to 6.2 ± 3.0 ng/ml/h at baseline and increased to 10.3 ± 10.2 ng/ml/h 90 min after captopril with a large scatter in the data (P 0.001). A positive response after captopril according to the test criteria was present in 62 patients. In this group plasma renin concentration increased from a baseline value of 9.2 ± 4.8 ng/ml/h to 28.8 ± 18.3 ng/ml/h after captopril (P 0.001). Further diagnostic workup by arteriography revealed significant renal artery stenosis (at least 50% narrowing of at least one renal artery) in 11 of these patients. In the 423 patients with a negative captopril response renal artery stenosis was found in three cases; arteriography, however, was not performed in all patients. With these limitations in mind sensitivity and specificity of the test were calculated as 79% and 89%, respectively. The prevalence of renal artery stenosis was 2.4% in patients with borderline and moderate hypertension compared to 11.8% in severely hypertensive patients. Blood pressure response after captopril Mean arterial pressure (1/3 (P sys P dia ) + P dia ) in the group with a positive response decreased from 111 ± 3to99 ± 3 mm Hg which was a significantly greater drop than in the group with a negative response (P 0.01) where the respective values were 114 ± 14 and 106 ± 16 mm Hg. Other diagnoses Clinical follow-up showed that 15 patients of the captopril-positive group were later normotensive and in one patient renal parenchymal hypertension was diagnosed. In the captopril-negative group seven patients were diagnosed having primary hyperaldosteronism and one patient had pheochromocytoma. In nine patients of this group hypertension was attributed to renal insufficiency. Normotensive blood pressure values were later obtained in 65 patients of this group. Renal function Twenty-one of the patients had a plasma creatinine value 1.5 mg/dl. The test was negative in 19 patients. In one of them renal artery stenosis was found. One patient of this group had a false positive captopril response. Complications No severe complication occurred during the test. Discussion Non-invasive tests for the detection of renovascular hypertension are needed since the prevalence of the disease in the general hypertensive population is low. In the past essentially three tests have been used as a screening test: renal scintigraphy following captopril administration (captopril renography), duplex scanning and the captopril test; each of them has its own advantages and disadvantages. The role of the captopril test for the diagnosis of renovascular hypertension has been controversial for several reasons: Renovascular hypertension vs renal artery stenosis A more general concern is that the term renal artery stenosis is not synonymous for renovascular hypertension. One has to take into consideration that a pure anatomical diagnosis of renal artery narrowing by whatever means does not necessarily prove the existence of renovascular hypertension. In other words the haemodynamic significance of a renal artery stenosis is hard to predict. Some authors require an angiographic stenosis of at least 70 to 75% to make the diagnosis, 6 8 whereas others use a cut-off of 50% narrowing. 9 Functional data, however, are usually not provided in most published series on any diagnostic procedure for the disease. With some limitations the most widely accepted functional criterion is cure or improvement (reduced number of antihypertensive medications) of hypertension after revascularisation, information that rarely is provided. Heterogeneity of the disease The picture of renal artery stenosis is extremely heterogeneous with regard to pathogenesis (atherosclerotic, fibromuscular etc), anatomy (unilaterlal, bilateral, solitary kidney, renal transplant), localisation (ostial, proximal, distal etc) and coexisting conditions (eg, renal insufficiency). It has been shown that the renin response after captopril

3 depends on these factors, eg, it is less evident in most cases with bilateral disease, in patients with renal insufficiency, younger age and in blacks. 3,10 Furthermore, ischaemic nephropathy with or without stenosis of the main renal artery or its branches has only recently been recognised as a clinical entity that shares functional properties with renovascular disease. Ischaemic nephropathy is often seen in diabetics and in patients with connective tissue disorders. 11 In these patients the renin response after captopril may be similar to patients with renovascular disease due to intrarenal small vessel abnormalities, even in the abscence of stenosis of the main renal artery or its branches. A higher prevalence of ischaemic nephropathy due to small vessel disease may therefore partly be responsible for a higher rate of false positive test results. Limited diagnostic accuracy There is a large variation in the published data on the diagnostic accuracy of the test, with results on sensitivity ranging from 40 to 100% and on specificity from 72 to 100%. 3 In the special group of renal transplant recipients a sensitivity of 75% and a specificity of 67% was found after 25 mg of captopril. 12 The prevalence of the disease in the cohort to be studied influences the diagnostic value of a test. 10,13 A higher rate of false positive results may therefore be obtained if patients with high renin essential hypertension with an exaggerated renin response after ACE inhibition are studied who per definition have a low probability for renovascular disease. 14 On the other hand when a hypertensive population with a high level of clinical suspicion for renovascular hypertension is investigated the captopril test does not discriminate appropriately between patients with essential and renovascular hypertension. 15 The prevalence of the disease in our own patients was only 3% which reflects the situation in the general hypertensive population probably quite well. Our study demonstrates that as can be expected almost 50 tests have to be performed to identify one patient with renal artery stenosis when a more general hypertensive population with a natural prevalence is investigated. Another reason for the divergent results may be the fact that most groups use their own protocol and that there is no generally accepted instruction on how to proceed. Harmonisation and standardisation, however, appear to be prerequisites for a meaningful interpretation of the results and a comparison between the various studies. Standardisation includes cessation of antihypertensive therapy, salt diet and food intake before the test, position during the testing and a valid renin assay. Table 1 attempts to give an overview on the studies that were published in the last decade (except renal transplant patients). It is obvious that various protocols and different diagnostic criteria are still being applied. Not surprisingly, similar to studies of the 1980s divergent results were obtained with respect to diagnostic accuracy of the test. If every group uses his own protocol, validation of the test criteria needs to be ensured by angiography-controlled prospective data rather than by post hoc adjustments, a criticism that also applies to our own study. It is hard to say whether such an expenditure is really justified or whether other ways can be found. Competing tests Other ways appear on the horizon since newer angiographic technologies that are only minimally invasive compete with the test. Spiral computerised tomography angiography (CTA) and magnetic resonance angiography (MRA) of the renal artery provide fascinating pictures and these technologies are currently being investigated in such patients. The introduction of paramagnetic contrast-media almost revolutionises the diagnostic reliability of MRA since now the problem of missed accessory renal arteries is mostly overcome. The current published reports suggest that a normal MRA is highly indicative for normal renal arteries, virtually abolishing the risk of false negative readings. 24 In addition, the method provides the possibility for the measurement of single kidney blood flow and filtration rate in the course of a single procedure. 25,26 A recent editorial comment concluded that it is very likely that MRA will replace all other diagnostic procedures for detection of renal artery stenosis. 27 In patients with normal renal function spiral CTA with intravenous contrast media appears to be a good alternative since the diagnostic accuracy is similar to MRA. 28,29 Due to the required high volume of contrast media it is usually not recommended for patients with renal function impairment. 27 The role of duplex scanning and captopril renography has extensively been discussed in previous publications. 3,10,30,31 Risks The captopril test has some risks which essentially are related to an overshooting blood pressure response. One report acclaims that 15% of the patients tested experienced reversible cerebral symptoms, 18 possibly due to the high prevalence of carotid artery disease in patients with atheromatous renal artery stenosis. 32 Other authors state that the test is well tolerated... and easy to perform. 7 Slightly anecdotal is the notion that it was occasionally necessary to place the patient in the supine position and to initiate a saline infusion. 2 It is obvious that the true rate of complications is hard to assess since the published data are not coherent in this regard and certainly far from complete. Also, the captopril doses used are not identical nor are other test variables (Table 1), eg, seated or supine position. Altogether, the test is certainly not without any risks. Costs Last but not least cost efficacy of a test should be taken into consideration. Blaufox and coworkers 33 performed a meta-analysis on various diagnostic tests comparing sensitivity and specificity, which were used to project cost per patient cured or 433

4 434 Table 1 Characteristics of studies for the use of captopril in the diagnosis of renovascular hypertension published in the last decade First author All pts/rvh pts Medication/mg Positive test criteria Arteriography- Sensitivity Specificity Prevalence (%) captopril for renin response confirmed (%) (%) Davidson 16 36/25 (69) discontinued/50 # post 5.7 ng/ml/h yes Degenhardt /24 (21) discontinued/0.5 mg/kg pre 2.0, post # yes ng/ml/h Frederickson 7 100/29 (29) discontinued/50 post 5.7 ng/ml/h yes Gosse 8 103/11 (10) partially maintained/50 + yes Hansen 18 47/11 (23) discontinued/ ## yes Idrissi 19 62/12 (19) discontinued/*** **yes Iwaoka /18 (9) discontinued/50 post 10.6 ng/ml/h yes Kutkuhn /44 (37) *maintained/25 posterior probability yes Postma 9 149/44 (30) discontinued/25 increase 16 ng/ml/h yes Schreij 15 46/20 (43) discontinued/25 ROC yes ROC ROC Svetkey 22 66/11 (17) discontinued/25 4 ng/ml/h **yes Takata 23 50/28 (56) discontinued/25 post 16 ng/ml/h yes *Except diuretics and ACE-inhibitors; **i.v. DSRA; ***sequential dosage, dependent on blood pressure response. # Post captopril value only available; # only when captopril test or i.v. DSA was positive; ## only when 131 -l-hippuran isotopic renography was positive; +various criteria according to Muller et al 2 ; ++increase not specified; only 2 days before test; receiver operator characteristic curves. improved for each modality. They found that the Doppler and the captopril test were less cost-effective than captopril renography and arteriography. To our knowledge no such comparative analysis exists for the new angiographic technologies. Presently, the cost of MRA is still too high to recommend it as a general screening method for hypertensive patients, restricting its use to patients with a high clinical suspicion for the disease. If a cost reduction is possible the outstanding diagnostic potential of the method is likely to pave the way for a broader application in patients with suspected renovascular hypertension. 27 Almost a decade ago a cautionary note concluded that the captopril test has a number of weaknesses; nevertheless, the commentators expressed their hope that the test should prove to be a very useful diagnostic tool in selected patients. 34 Approaching the year 2000 a critical evidence-based analysis in our opinion would rather suggest that for reasons exclaimed above, the test is not accurate and not simple enough to play an important role in the diagnostic work-up of even selected patients with suspected renovascular disease in the future. We believe its time is over since better tools have become available. Acknowledgements This work was supported by the Dr Bodo Sponholz Stiftung, Frankfurt am Main. The authors thank Mrs Astrid Mösbauer for measuring the plasma renin concentration. Parts of this study were presented in abstract form at the Annual Meeting of the German Hypertension League (Dreiländertreffen), Freiburg, Germany, References 1 Case DB, Laragh JH. Reactive hyperreninemia in renovascular hypertension after angiotensin blockade with saralasin or converting enzyme inhibitor. Arch Intern Med 1979; 91: Muller FB et al. The captopril test for identifying renovascular disease in hypertensive patients. Am J Med 1986; 80: Pickering TG, Mann SJ. Renovascular hypertension: medical evaluation and nonsurgical treatment. In: Laragh JH, Brenner BM (eds). Hypertension: Pathophysiology, Diagnosis and Management. Raven: New York, 1995, pp Gaul MK, Linn WD, Mulrow CD. Captopril-stimulated renin secretion in the diagnosis of rnovascular hypertension. Am J Hypertens 1989; 2: Boucher R, Veyrat R, de Champlain J, Genest J. New procedures for measurement of human plasma angiotensin and renin activity levels. Canad Med Ass J 1964; 90: Pickering TG et al. Predictive value and changes of renin secretion in hypertensive patients with unilateral renovascular disease undergoing successful renal angioplasty. Am J Med 1985; 76: Frederickson ED et al. A prospective evaluation of a simplified captopril test for the detection of renovascular hypertension. Arch Intern Med 1990; 150: Gosse P et al. Captopril test in the detection of renovascular hypertension in a population with low prevalence of the disease. A prospective study. Am J Hypertens 1989; 2: Postma CT et al. The captopril test in the detection of renovascular disease in hypertensive patients. Arch Intern Med 1990; 150: Mann SJ, Pickering T. Detection of renovascular hypertension. Ann Intern Med 1992; 117: Sos TA, Trost DW. Renal vascular disease as a cause of hypertension. Curr Opin Nephrol Hypertens 1995; 4: Erley CM et al. Noninvasive procedures for diagnosis of renovascular hypertension in renal transplant recipients a prospective analysis. Transplantation 1992; 54: Pickering TG. Diagnosis and valuation of renovascular hypertension. Circulation 1991; 83 (Suppl I): I-147 I Gerber LM et al. Response to the captopril test is dependent on baseline renin profile. J Hypertens 1994; 12: Schreij G et al. Captopril test with blood pressure and peripheral renin as response variables in hypertensive

5 patients with suspected renal artery stenosis. J Hum Hypertens 1995; 9: Davidson RA, Barri YM, Wilcox CS. The simplified captopril test: an effective tool to diagnose renovascular hypertension. Am J Kidney Dis 1994; 24: Degenhardt S et al. The value of the captopril test in the work-up of hypertension. Klin Wochenschr 1989; 67: Hansen PB, Garsdal P, Fruergaard P. The captopril test for identification of renovascular disease: value and immediate adverse effects. J Intern Med 1990; 228: Idrissi A et al. The captopril challenge test as a screening test for renovascular hypertension. Kidney Int 1988; 34 (Suppl 25): S138 S Iwaoka T et al. The usefulness of the captopril test as a simultaneous screening for primary aldosteronism and renovascular hypertension. Am J Hypertens 1993; 6: Kutkuhn B et al. Validity of the captopril test for identifying correctable unilateral renovascular hypertension. Clin Exp Hyper A 1991; A13: Svetkey LP et al. Prospective analysis of strategies for diagnosing renovascular hypertension. Hypertension 1989; 14: Takata M et al. Diagnostic value of captopril test in hypertensive patients with renal artery stenosis. Angiology 1994; 45: Cobelli FD et al. Renal artery stenosis: value and screening with three-dimensional phase-contrast MR angiography with a phased-array multicoil. Radiology 1996; 201: Wolf RL et al. Measurement of normal renal blood flow: cine phase-contrast MR imaging vs clearance of p-aminohippurate. Am J Roentgenol 1993; 161: Ros PR et al. Diagnosis of renal artery stenosis: feasibility of combining MR angiography, MR renography, and gadopentate-based measurements of glomerular filtration rate. Am J Roentgenol 1995; 165: Olbricht CJ, Arlart IP. (Editorial comment). Magnetic resonance angiography the procedure of choice to diagnose renal artery stenosis? Nephrol Dial Transplant 1998; 13: Galanski M et al. Renal artery stenoses: spiral CT angiography. Radiology 1993; 189: Olbricht CJ et al. Minimally invasive diagnosis of renal artery stenosis by spiral computed tomography angiography. Kidney Int 1995; 48: Mann SJ et al. Captopril renography in the diagnosis of renal artery stenosis. Accuracy and limitations. Am J Med 1991; 90: Spies KP et al. Color-coded duplex sonography for non-invasive diagnosis and grading of renal artery stenosis. Am J Hypertens 1995; 8: Missouris CG et al. High prevalence of carotid artery disease in patients with atheromatous renal artery stenosis. Nephrol Dial Transplant 1998; 13: Blaufox MD, Middleton ML, Bongiovanni J, Davis BR. Cost efficacy of the diagnosis and therapy of renovascular hypertension. J Nucl Med 1996; 37: McCarthy JE, Weder AB. (Editorial). The captopril test and renovascular hypertension. A cautionary tale. Arch Intern Med 1990; 150: