Colour Doppler ultrasound in renal transplant artery stenosis: Which Doppler index?

Clinical Radiology (1995) 50, 618 622

Colour Doppler Ultrasound in Renal Transplant Artery Stenosis: Which Doppler Index? G. M. BAXTER, H. IRELAND, J. G. MOSS, P. N. HARDEN*, B. J. R. JUNOR*, R. S. C. RODGER* and J. D. BRIGGS*

*Department of Radiology & Renal Medicine, Western Infirmary NHS Trust, Glasgow, UK A prospective study comparing colour Doppler ultrasound (US) with the 'gold standard' of intra-arterial digital subtraction angiography in the evaluation of renal transplant artery stenosis was performed. Both the intrarenai vessels and the transplant renal artery were examined by Doppler US. Diagnostic arteriography was performed only if, on Doppler, the peak systolic velocity in the transplant renal artery exceeded 1.5 ms -1. Of 109 patients, the transplant artery could not be visualized using colour Doppler US in three, and these were omitted from statistical analysis. Of the remaining 106 patients, 31 had a peak systolic velocity greater than 1.5 ms-~ in the transplant renal artery and were referred for DSA. Of the multiple renal Doppler indices recorded, the peak systolic velocity in the transplant artery was the best discriminating measurement for the detection of renal artery stenosis. A peak systolic velocity of >2.5 ms-' in the transplant renal artery had a sensitivity of 100% and a specificity of 95% for the detection of renal artery stenosis (> 50% diameter reduction). Although a significant difference in Pulsatifity Index, Resistive Index, Acceleration Index and Acceleration Time was recorded from the intrarenal vessels in the angiographically normal and stenosed groups with Doppler, these measurements were less useful as discriminating diagnostic tests. In conclusion, the peak systolic velocity in the transplant renal artery is the most sensitive Doppler criterion for renal artery stenosis and is sensitive and specific enough to be used as a screening test. The intrarenal acceleration time and index should not be used in isolation. Baxter, G.M., Ireland, H., Moss, J.G., Harden, P.N., Junor, B.J.R., Rodger, R.S.C. & Briggs, J.D. (1995). Clinical Radiology 50, 618-622. Colour Doppler Ultrasound in Renal Transplant Artery Stenosis: Which Doppler Index?

Accepted for Publication 5 January 1995

Despite the relative shortage of donor kidneys there has been a steady growth in the number of renal tralasplants performed in the UK since the 1970s. With improved surgical techniques and better immunosuppression both public and medical expectations of a successful transplant are high and as many as 90% of transplant kidneys will be functioning at one year, with a 5 year graft survival in the region of 65% or more [1]. With improved outcome and graft longevity, secondary vascular complications such as transplant artery stenosis are well recognized. Renal artery stenosis (RAS) has been reported to occur in 3% to 15% of transplant kidneys and usually presents within three years of transplantation [2]. Clinical presentation is variable and includes deteriorating renal function and resistant hypertension. Clinical findings such as graft bruits are both uncommon and unreliable. In addition, exclusion of renal artery stenosis (RAS) before the use of angiotensin converting enzyme (ACE) inhibitors would be advantageous as deterioration in renal function on starting ACE inhibitors is well recognized in the presence of a significant RAS. Diagnostic arteriography remains the 'gold standard' but is invasive, expensive and involves the use of contrast Correspondence to: Dr G. M. Baxter, Department of Radiology, Western Infirmary/GGH University NHS Trust, Glasgow G11 6NT, UK. 9 1995 The Royal College of Radiologists.

agents which may rarely initiate an allergic reaction. In view of these disadvantages there has been a search for a suitable non invasive alternative diagnostic technique. Captopril renography has met with mixed success and when compared with intraarterial renal angiography has a sensitivity and specificity of only 75% and 84%, respectively [3]. Magnetic Resonance imaging (MRI) using phase contrast techniques has shown a better sensitivity and specificity of 83% and 97%, respectively, in the diagnosis of transplant artery stenosis [4] but it is relatively expensive, and not yet widely available in the UK. Doppler ultrasound (US) has met with mixed success [5-8] in the diagnosis of RAS in both the native and transplant kidney. As direct imaging of the artery is required, the examination can be technically challenging and time-consuming. Recently, a phenomenon known as the 'parvus tardus' effect has been described in the intrarenal waveforms downstream from a proximal RAS. Examination to detect this effect is easier technically. However, although in the native kidney, initial results were encouraging [9], the report of a subsequent study has expressed less enthusiasm [10]. With regard to the transplant kidney, a wide range of diagnostic Doppler criteria exist [11-13]. There is no agreement on a specific diagnostic cut-off value for renal artery stenosis. Also, the accuracy of the 'parvus tardus' effect has not been prospectively assessed. We have performed a prospective study which

COLOUR DOPPLER ULTRASOUND IN RENAL TRANSPLANT ARTERY STENOS1S

compares the accuracy of colour Doppler US with the gold standard of intra arterial Digital subtraction angiography in patients with suspected renal transplant artery stenosis. In addition, using colour Doppler, we have compared the peak systolic velocity in the transplant renal artery with the intrarenal waveform measurements of Acceleration Index and Time, to evaluate which was superior in the diagnosis of RAS. PATIENTS AND METHODS Patients in whom a transplant RAS was suspected were referred for colour Doppler US. Referral criteria included hypertension, the presence of a graft bruit on auscultation, unexplained deterioration in renal function or exclusion of a significant stenosis before ACE inhibitor therapy. Colour Doppler examination was performed with a 3.5MHz vector probe (Acuson, Mountainview, CA, USA) and the examination time was restricted to a maximum of 20 rain. Real time US was initially performed to assess both kidney size and appearance and exclude hydronephrosis. A colour flow examination of the intrarenal vessels was then carried out and arterial Doppler spectra sampled from the upper, mid- and lower-pole interlobar arteries of the kidney. Six Doppler indices were recorded from each waveform, namely peak systolic velocity, end diastolic velocity, Resistive Index (RI), Pulsatility Index (PI), Acceleration Index (AI) and Acceleration Time (AT). A mean of three measurements for each individual Doppler index was calculated for each kidney. The adjacent iliac artery was then visualized to exclude a significant iliac stenosis. Finally the transplant renal artery was examined from its origin at the iliac artery to the renal hilum. Doppler spectral analysis was performed along the length of the artery. A peak systolic velocity of _~ -1 1.5 ms was considered normal and no further investigation was performed. The presence or absence of turbulence was noted in the transplant artery and graded as mild, moderate or severe. Digital subtraction angiography was performed on all patients who had a peak systolic velocity of >l.Sms -~ in the transplant renal artery. Digital subtraction angiography was performed as an out patient procedure with a 3-4 h hospital stay. Vascular access was obtained through the ipsilateral groin to the transplant kidney using a 3-F pigtail catheter. A flush angiogram of the lower aorta and iliac vessels was initially performed followed by a minimum of three projections of the transplant artery. Selective catheterisation of the transplant renal artery was only undertaken when this was unsatisfactory. All angiograms were examined by a vascular radiologist and stenoses were graded by comparison of the diameter of the narrowed segment to that of the more proximal artery. All angiograms were performed within one month of the Doppler examination. Clinical details including renal function, blood pressure and the number of previous rejection episodes were recorded for each patient. Statistical analysis was initially performed using the Kruskal Wallis test followed by the Mann-Whitney test for both intrarenal and transplant artery Doppler measurements. Analysis of the transplant artery peak systolic velocity as a continuous variable was performed using Spearman

619

Rank Correlation. All statistically significant results (P < 0.05) were Bonferroni corrected. RESULTS A total of 109 renal transplant patients were scanned with colour Doppler US. In three, the transplant artery could not be visualized due to obesity, bowel gas or both. Of the remaining 106, 31 had a peak systolic velocity in the transplant artery of > 1.5 ms-~ on Doppler examination and proceeded to diagnostic angiography. The median time from transplantation was 3.3 years (range 7 months to 9 years, 7 months). Of the 31 patients who had both examinations (M : F 21 : 10), angiography confirmed the presence of a significant stenosis (>50% diameter narrowing) in 10. All 10

(a)

9

9 1995 The Royal College of Radiologists, Clinical Radiology, 50, 618-622.

(b) Fig. 1 - (a) Doppler US through the proximal transplant renal artery. Spectral waveform analysis shows a peak systolic velocity of 4.51 ms-', indicative of a significant stenosis. (b) Intra arterial Digital Subtraction arteriogram of the same patient as in (a) confirming the above Doppler findings. A significant stenosis is present in the proximal transplant artery (arrow).

620

CLINICAL RADIOLOGY

Table 1 - Diagnostic accuracy of Doppler measurements in renal transplant artery stenosis Doppler measurement

SENS

SPEC

PPV

NPV

ACC

PSV Transplant artery >2.5ms -I Intrarenal AI < 1.5 ms -2 Intrarenal AT >0.08 s

100

95

91

i00

97

86

87

75

93

86

100

73

64

100

82

Number of patients = 31. PSV, peak systolic velocity; AI, acceleration index; AT, acceleration time; SENS, sensitivity; SPEC, specificity; PPV, positive predictive value; NPV, negative predictive value; ACC, accuracy.

patients had a peak systolic velocity in the transplant artery of >2.5 ms -1 (range 2.5-4.47 ms -1) (Fig. la, b). Of the 21 patients with a normal angiogram, 20 had a peak systolic velocity in the transplant artery of <2.5ms -I. The sensitivity and specificity of a peak systolic velocity of >_2.5ms -1 in the transplant artery for RAS were thus 100% and 95%, respectively (Table 1). Significant differences were observed in the intrarenal PI, RI, AI and AT between the angiographically stenosed and the angiographically normal group (Table 2). No differences in intrarenal peak systolic or end diastolic velocities were detected. No difference in intrarenal Doppler indices was detected between the angiographically normal group and those in whom diagnostic arteriography was not performed (Table 2). As a diagnostic test for RAS the intrarenal AI and AT (using cut-off values of 1.5 ms -2 and 0.08 s, respectively) compared poorly with the transplant artery peak systolic velocity (Table 1). Using Spearman Rank Correlation, a relationship was demonstrated between the transplant artery peak systolic velocity and the intrarenal Doppler measurements. As the peak systolic velocity in the transplant renal artery increased the RI, PI and AI fell, whilst the AT increased (Fig. 2a, b, c). In addition, high peak systolic velocities, and thus stenoses, were associated with the most severe forms of arterial turbulence. There was no difference in patient age, kidney size, serum creatinine or blood pressure between the angiographically proven RAS and normal groups. A difference in the number of rejection episodes was noted between the RAS (55% had one or more episodes of rejection) and normal groups (25% had one or more episodes of rejection) although this did

not quite reach statistical significance (P = 0.07, Fisher's Exact Test). Twenty-nine of 31 patients who had angiography, had cadaveric transplants, and all had end to side surgical anatomoses to the external iliac artery, whilst two live donor transplants had end to end anastomoses to the internal iliac artery. The transplant kidney was supplied by an accessory renal artery in six patients, and in two of those patients, a significant stenosis of one was present. US detected both these stenoses but failed to detect the accessory renal vessel in all cases. Of the 75 patients who did not have angiography (peak systolic velocity transplant artery
Table 2 -Intrarenal Doppler arterial indices in normal and stenosed transplant renal arteries (median and inter quartile range) Intrarenal Doppler index

No angio P S V <1.5 ms -1 (n = 75)

Angio normal (n = 21)

Angio, R A S (n = 10)

Kruskal Wallis test

PSV EDV RI PI AI AT

21 (18-26.2) 7 (5-9) 67 (63-71.2) 126 (114-147.2) 2.37 (1.75-3.42) 0.059 (0.045-0.08)

22.5 (19.5-27.2) 7.5 (6-10.2) 66.5 (62-68.5) 122.5 (105.5-134) 1.8 (1.58-2.21) 0.072 (0.061-0.085)

19.0 (14-21) 8 (6-9)

P P P P P P

57 (55-65)

94.5 (77-102) 0.81 (0.72-1.3) 0.101 (0.091-0.152)

= = = < < <

0.398 0.288 0.004 0.001 0.001 0.001

psv, Doppler peak systolic velocity transplant renal artery. PSV, peak systolic velocity; EDV, end diastolic velocity; RI, resistive index; PI, pulsatility index; AI, acceleration index (ms-Z); AT, acceleration time (s). Using the Mann Whitney Test, significant differences were detected for RI, PI, AI and AT between the angiographically normal and stenotic groups and between the stenotic group and those who did not have angiography (P < 0.05, P < 0.01). No significant difference was recorded between those patients with angiographically normal transplant arteries and those with a psv
COLOUR DOPPLER ULTRASOUND IN RENAL TRANSPLANT ARTERY STENOSIS

(a)

(b)

(c) Fig. 2 - (a) Digital arteriogram demonstrating back to back high grade stenoses at the origin of the transplant renal artery (arrows). (b) Doppler US of the same patient as in (a). A high velocity jet is present in the renal artery with a peak systolic velocity of 3.73 ms -l indicating transplant artery stenosis, Marked spectral broadening is also present. (c) Doppler interrogation of an intrarenal artery downstream from the stenotic jet demonstrated in (b). The waveform is flattened with loss of its normal sharp systolic upstroke. The appearances are those of the 'parvus tardus' effect with prolongation of the acceleration time and reduction in acceleration index.

9 1995 The Royal College of Radiologists, Clinical Radiology, 50, 618-622.

621

(P = 0.07) but this may in part reflect the relatively small numbers. Another factor to be considered is the potential hazard of acute deterioration in graft function with ACE inhibitor therapy in patients with RAS. An accurate diagnostic test is thus required and, to date, this has meant digital angiography, which is invasive with a recognized morbidity and mortality. In view of these disadvantages there has been a search for a suitable non invasive alternative. One such method is captopril renography using 99mTc-DTPA before and after ACE inhibition with measurement of both glomerular filtration rate and effective renal plasma flow, pre- and post-ACE administration. However, this test is time consuming and only has a sensitivity and specificity of 75% and 84%, respectively [3]. The possible use of Doppler and colour Doppler US as a replacement for angiography was initially assessed in native reno-vascular disease. Results vary widely between different investigative groups with reported sensitivities of between 0% to 95% [5,6]. Accessory renal vessels cannot be detected and disappointingly, the addition of colour flow imaging has had no impact on these results [5,8]. Numerous diagnostic criteria have been used and include the peak systolic velocity in the renal artery, the ratio of the peak systolic velocity in the renal artery to aorta (RAR) and a PI difference of greater than 0.12 between kidneys [23], but none of these have been found to be consistently reliable. Adequate Doppler interrogation of the renal artery is difficult, and most of all, extremely time consuming [8]. As a result, attention has turned to the intra renal vessels and the downstream Doppler changes that occur secondary to renal artery stenoses, the so called 'parvus-tardus effect.' This results in a more rounded, smaller amplitude Doppler waveform with a consequent reduction in AI and increase in AT. Initial results were extremely encouraging [9] with sensitivities of 95% recorded. Recently conducted trials, however, have been less enthusiastic [10] with significant differences only recorded between normal and severely stenotic arteries (>80%). The parvus tardus effect was initially thought to occur secondary to a pressure drop over the stenosis [24]. However, recent work has suggested that it is the compliance of the post stenotic vessel in combination with the stenosis that produces the abnormal waveform pattern [25]. These studies have largely been confined to the native kidneys, although transplant kidneys have been included in some. Renal transplants should be considered separately from native kidneys as they differ in anatomical .site and local flow haemodynamics and are subject to vascular and parenchymal insults peculiar to transplantation. In renal transplantation, as in the native kidney, both results and diagnostic criteria in RAS vary considerably [11,12] and this, in combination with the unsatisfactory experience of colour Doppler in native RAS, has contributed to scepticism about the value of the technique in transplant RAS. Our results have shown that although there are changes in the intrarenal Doppler indices in transplant artery stenosis, these, in isolation, are not good enough to lead to an accurate diagnosis. The most reliable diagnostic measurement was the peak systolic velocity in the transplant renal artery with a sensitivity of 100% and a specificity of 95% for RAS. Although the cut-off value of >2.5 ms -1 is significantly

622

CLINICAL RADIOLOGY

higher than that quoted in many studies [11] we have found this measurement to carry a very satisfactory degree of accuracy. In addition, the cut-off values of 1.5ms -2 and 0.08s for AI and AT respectively, are different from those quoted for the native kidney [9,10] and this presumably reflects the unique situation of the transplant kidney. Exact diagnostic cut-off values cannot be entirely satisfactory in renal transplantation due to the many inherent imaging variables such as patient build, vascular anatomy and vessel tortuousity. At our centre, all cadaveric renal transplant arteries are surgically anastomosed end to side to the external iliac artery when possible and vessel tortuosity presents a challenge to the sonologist, as accurate angle correction is imperative. This may account, in part, for the wide range of cut-off values in the literature. Live donor transplant arteries are, however, routinely anastomosed end to end to the internal iliac artery. The transplant artery is technically more difficult to image with Doppler US due to the depth of the vessel, and this could again contribute to the variation in results, as the number of live donor grafts vary between centres. Our results suggest that a transplant artery peak systolic velocity value of _>2.5ms -1 is a suitable diagnostic cut-off value for RAS. It is difficult to be dogmatic about such values, as a large proportion of these patients with normal transplant artery Doppler velocities ( 1.5 and <2.5 ms -1 had normal arteriograms, and indeed that 23 of 24 patients with transplant artery velocities of
4 5 6 7 8 9 10 11

12 13 14 15 16

17

18 19 20 21

Acknowledgements.The authors would like to thank Dr J. G. Love and Dr G. Murray for their help and assistance with the statistical analysis for this paper.

22 23

REFERENCES ! Merion RM, White DJG, Thiru S e t aL Cyclosporin: five years experience in cadaveric renal transplantation. New England Journal of Medieine 1984;310:148-154. 2 Gedroyc WM, Reidy JF, Saxton HM. Arteriography of renal transplantation. Clinical Radiology 1987;28:239-243. 3 Erley CM, Duda SH, Wakat JP et aL Noninvasive procedures for

24 25

diagnosis of renovascular hypertension in renal transplant recipients - a prospective analysis. Transplantation 1992;54:863 867. Gedroyc WM, Negus R, al Kutoubi A et al. Magnetic resonance angiography of renal transplants. Lancet 1992;339:789-791. Berland LL, Koslin DB, Routh WD et al. Renal artery stenosis: prospective evaluation of diagnosis with color duplex US compared with angiography. Radiology 1990;174:421-423. Desberg AL, Paushter DM, Lammert GK et al. Renal artery stenosis; evaluation with color Doppler flow imaging. Radiology 1990;177:749-753. Hansen K J, Tribble RW, Reavis SW et al. Renal duplex sonography: evaluation of clinical utility. Journal of Vascular Surgery 1990;12:227 236. Postma CT, van Aalen J, de Boo T et al. Doppler ultrasound scanning in the detection of renal artery stenosis in hypertensive patients. British Journal of Radiology 1992;65:857-860. Stavros AT, Parker SH, Yakes WF et al. Segmental stenosis of the renal artery: pattern recognition of tardus and parvus abnormalities with Duplex sonography. Radiology 1992;184:487-492. Kliewer MA, Tupler RH, Carroll BA et al. Renal artery stenosis: analysis of Doppler waveform parameters and tardus-parvus pattern. Radiology 1993;189:779-787. Duda SH, Erley CM, Wakat JP et al. Post transplant renal artery stenosis outpatient intraarterial DSA versus color aided duplex Doppler sonography. European Journal o f Radiology 1993;16: 95-101. Snider JF, Hunter DW, Moradian GP et al. Transplant renal artery stenosis: evaluation with duplex sonography. Radiology 1989;172: 1027-1030. Taylor KJW, Morse SS, Rigsby CM et al. Vascular complications in renal allografts: detection with duplex Doppler US. Radiology 1987;162:31-38. Rifkin MD, Needleman L, Pasto ME et al. Evaluation of renal transplant rejection by duplex Doppler examination: value of the resistive index. American Journal of Radiology 1987; 148:759-762. Rigsby CM, Taylor KJW, Weltin G e t al. Renal allografts in acute rejection: evaluation using duplex sonography. Radiology 1986; 158:375-378. Genkins SM, Sanfilippo FP, Carroll BA. Duplex Doppler sonography of renal transplants: lack of sensitivity and specificity in establishing pathologic diagnosis. American Journal of Radiology 1989;152:535-539. Kelzc F, Pozniak MA, Pirsch JD et al. Pyramidal appearance and resistive index: insensitive and nonspecific indicators of acute renal transplant rejection. American Journal of Radiology 1990;155: 531 535. Baxter GM, Morley P, Dall B. Acute renal vein thrombosis in renal allografts: new Doppler ultrasonic findings. Clinical Radiology 1991;43:125-127. Reuther G, Wanjura D, Bauer H. Acute renal vein thrombosis in renal allografts: detection with duplex Doppler ultrasound. Radiology 1989;170:557 558. Renowden SA, Blethyn J, Cochlin DL. Duplex and colour flow sonography in the diagnosis of post biopsy arteriovenous fistulae in the transplant kidney. Clinical Radiology 1992;45:233-237. Merkus JWS, Zeebregts CJAM, Hoitsma AJ et al. High incidence of arteriovenous fistula after biopsy of kidney allografts. British Journal of Surgery 1993;80:310-312. Lacombe M. Arterial stenosis complicating renal allotransplantation in man. Annals of Surgery 1975;181:293. Bardelli M, Jensen G, Volkmann R et al. Non invasive ultrasound assessment of renal artery stenosis by means of the gosling pulstility index. Journal of Hypertension 1992;10:985-989. Lafortune M, Patriquin H, Demeule E et al. Renal arterial stenosis: slowed systole in the downstream circulation - experimental study in dogs. Radiology 1992; 184:475-478. Bude RO, Rubin JM, Platt JF et al. Pulses tardus: its cause and potential limitations in detection of arterial stenosis. Radiology 1994;190:779-784.

9 1995 The Royal College of Radiologists, ClinicalRadiology, 50, 618-622.