Ian R. Davidson

Diagnosis of Lower-Limb Deep Venous Thrombosis: A Prospective Blinded Study of Magnetic Resonance Direct Thrombus Imaging

Despite considerable recent advances in diagnostic techniques for lower-limb deep venous thrombosis (DVT), current methods have disadvantages. Ultrasonography, the most accurate noninvasive test, is widely available and cheap. As such, it has largely replaced venography as the test of first choice for symptomatic DVT. In a recent meta-analysis, the sensitivity of ultrasonography was 89% overall for symptomatic DVT and 97% for above-knee thrombosis (1). Large outcome studies have shown that patients may be safely left untreated after a negative result on ultrasonography if they have a low clinical risk score, a low d-dimer level, or a negative result on repeated ultrasonography at 1 week (2-4). However, these strategies may be complex and still require 3% to 34% of outpatients and most inpatients to undergo repeated ultrasonography at 1 week (2-4). In practice, retesting after 1 week is inconvenient, and physicians often rely on a single test or request immediate venography (5). Other problems with ultrasonography include poor sensitivity for asymptomatic disease, difficulties in diagnosing DVT recurrence, and limited visualization in the pelvis (1, 6, 7). Impedance plethysmography is also commonly used; however, it has a lower diagnostic accuracy than ultrasonography and has similar weaknesses in the setting of recurrent thrombosis, asymptomatic DVT, and DVT below the knee or in the pelvis (1, 4, 6). Computed tomography and magnetic resonance imaging techniques can visualize DVT above the knee and in the pelvis but in general are unsuccessful below the knee (8-10). The ability of these techniques to diagnose DVT recurrence and asymptomatic disease has not been tested. Venography is the reference standard diagnostic test, but it has in large part been replaced by noninvasive tests. In clinical practice, it is the most reliable test for the diagnosis of asymptomatic thrombosis and thrombosis isolated within the calf or pelvis. However, imaging in the pelvis is inadequate in up to 24% of normal studies, and the proximal extent of thrombosis is frequently not delineated in patients with above-knee DVT (11). Underfilling of vessels and vessels overlying one another also create problems with venography below the knee. Studies have shown that interobserver variability for venography is high (10% to 16%), especially below the knee ( = 0.46 to 0.73 below the knee and 0.46 to 0.84 above the knee) (12, 13). In addition, a high proportion of studies are nondiagnostic for possible DVT recurrence (1, 6). A noninvasive test is needed that accurately diagnoses above-knee DVT and thrombus below the knee, in the pelvis, and in asymptomatic limbs. Unlike most imaging techniques, which identify thrombus as filling defects, magnetic resonance direct thrombus imaging (MRDTI) visualizes thrombus against a suppressed background (14). In an unblinded comparison with venography, we previously showed that MRDTI precisely visualizes acute deep venous thrombus (14, 15). In the current study, we sought to assess prospectively whether MRDTI is a reliable diagnostic test for suspected acute symptomatic DVT. Methods The ethics committee at our institution granted approval for the study, and all participants gave written informed consent. With the exceptions of pregnant women, patients with known contrast allergy, and those with renal failure, all patients with DVT suspected on the basis of lower limb symptoms are investigated by using venography at our institution. Participants were recruited after routine venography was done between May 1998 and September 1999. During this time, 338 consecutive patients underwent routine contrast venography. Consecutive patients with positive venograms were selected, along with one quarter of those with negative venograms, according to a predetermined random sequence. This protocol was chosen to equalize the numbers of positive and negative cases and was based on a 6-month audit of venograms in our institution that found that 22% of venograms were positive. Clinical diagnostic criteria were not used, and the decision to request investigation for suspected DVT had been made by the attending clinician; however, patients who did not have leg symptoms were not recruited. Other exclusion criteria were failed or inconclusive venography, failed or inconclusive MRDTI, contraindications to MRI, and claustrophobia (Figure 1). Individual venous segments that were nondiagnostic at venography were also excluded from analysis. Figure 1. Outline of the study. Magnetic resonance direct thrombus imaging was performed on all patients recruited within 48 hours of venography. The scans were interpreted by an experienced radiologist (reviewer A) and by a nonradiologist (reviewer B) trained to read MRDTI scans. For venograms and MRDTI scans, the reviewers noted the presence or absence of DVT; the diagnostic classification of DVT, divided into isolated calf DVT, femoropopliteal DVT, and ileofemoral DVT; and the presence of thrombus in the calf, femoropopliteal, and iliac venous segments. Venograms were obtained and initially reported by the radiologists on duty. This initial report was used to make recruitment decisions; if the results were discordant with those of MRDTI, ultrasonography was also performed. However, ultrasonography was not used in the calculations of the accuracy of MRDTI. After completion of the study, venograms were interpreted by an independent radiologist, and these results were used as the gold standard against which MRDTI was compared. Results of MRDTI and venography were reported without knowledge of the results of other tests and the other readings. The d-dimer level was measured in all patients at the time of the MRDTI scan by using the Nycocard (Nycomed Pharma AS, Asker, Norway) technique (normal level < 0.3 mg/L). Venography Venography was performed by cannulating a dorsal pedal vein with a 21-gauge needle and rapidly injecting 50 to 100 mL of iodinated contrast medium (I2, 300 mg/mL), with the patient supine and tilted 30 degrees with his or her feet downward. A tourniquet was applied above the ankle. Anteroposterior and two oblique views of the deep calf and popliteal veins were obtained. Views of the femoral and iliac veins were then obtained. The study result was considered positive if intraluminal filling defects were seen or persistent nonfilling of veins with a sharp cut-off was detected. Magnetic Resonance Imaging Magnetic resonance imaging was performed by using a 1.5-Tesla unit (Siemens Vision, Erlangen, Germany) with a T1-weighted magnetization-prepared three-dimensional gradient-echo sequence. The sequence included a water-only excitation radiofrequency pulse to abolish the fat signal, and the effective inversion time was chosen to nullify the blood signal. Imaging was performed from the ankle to the inferior vena cava in two imaging blocks with a total acquisition time of 12 minutes by using a 55-cm body coil. Both legs were imaged simultaneously. Scanning was performed by radiographers in all cases. Image assessment involved reading of coronal source data and standard image reconstruction techniques. Acute thrombus was diagnosed on the basis of its high signal against the suppressed background (Figure 2). Figure 2. Magnetic resonance direct thrombus imaging in three patients. A. arrows B. arrows C. single arrows double arrow Ultrasonography Color flow and compression ultrasonographic images of the symptomatic limb from the common femoral vein distally were obtained by using a 5-MHz linear array transducer. As much of the superficial femoral vein as possible was imaged, together with the popliteal vein and the calf veins. Augmentation of flow was used to verify patency. Examinations were performed by senior radiologists, and DVT was confirmed in all cases by noncompressibility on gray-scale images. The sonographer was unaware of the other test results, but in cases of possible isolated calf thrombosis, he or she was told to concentrate the examination below the knee to maximize accuracy in this region. Statistical Analysis Sensitivity and specificity were calculated for the overall diagnosis of DVT; diagnosis of isolated calf DVT, femoropopliteal DVT, and ileofemoral DVT; and presence of thrombus in the calf, femoropopliteal vein, and iliac vein. Exact CIs were calculated. Interobserver error was calculated for these observations by using the weighted statistic with equally spaced weights for positive, nondiagnostic, and negative studies. Confidence intervals for the statistic were calculated from asymptotic estimations of the standard error. Calculations were performed by using SPSS software (SPSS, Inc., Chicago, Illinois). Results One hundred four patients were recruited according to our protocol (Figure 1). The time between venography and MRDTI was less than 8 hours in 28 patients, 8 to 24 hours in 44 patients, and 24 to 48 hours in 32 patients. Age ranged from 20 to 95 years, and symptom onset varied from 1 to 35 days. Ninety-five patients were referred from medical specialties and 9 from surgical specialties; 47 were inpatients and 57 were outpatients. Both reviewers reported that 3 of 5 patients with ipsilateral total hip replacements had nondiagnostic MRDTI scans. Venography diagnosed femoropopliteal DVT in 1 of these patients and was negative in 2 patients. These 3 patients were excluded from further analysis, leaving 101 patients in the study. One patient could tolerate only the first scanning block from ankle to thigh level owing to claustrophobia; however, femoropopliteal DVT could still be diagnosed. All other patients tolerated MRI. Eighteen of 148 patients (12%) were excluded from the study. Fifteen patients could not undergo MRI because of contraindications (9 patients) or claustrophobia (6 patients), and 3 patients had inconclusive results on MRDTI. Venography failed (29 patients) or was inconclusive (11 patients) in 12% of patients (40 of 338). Venography was inconclusive

Anatomical suitability of ruptured abdominal aortic aneurysms for endovascular repair.

Purpose: To assess the anatomical suitability of ruptured abdominal aortic aneurysms (AAA) for emergency endovascular repair. Methods: All cases (46 patients [35 men; mean age 74 years, range 54–85]) in which computed tomographic angiography (CTA) confirmed AAA rupture over a 5-year period at our university hospital were reviewed for anatomical suitability for endovascular repair. Measurements were made by a radiologist experienced in anatomical assessment of CT criteria for elective endovascular aneurysm repair (EVAR). Results: The mean aneurysm neck length was 18 mm (range 0–59); 17 were conical, 13 straight, 4 barrel, and 6 reverse conical. Six cases had no proximal neck. Overall, 37 (80%) patients were unsuitable for EVAR according to our criteria. Nearly half the patients (22, 48%) had ≥2 adverse features. Unsuitable neck morphology (35, 76%) was the primary reason for exclusion, but CIA aneurysm (10, 22%) and EIA tortuosity (7, 15%) were secondary adverse features. Conclusions: With current stent-graft design, the majority of ruptured abdominal aortic aneurysms are anatomically unsuitable for endovascular repair.

A study on the patency of the inferior mesenteric and lumbar arteries in the incidence of endoleak following endovascular repair of infra-renal aortic enerysms

OBJECTIVE An endoleak is defined as the presence of contrast medium within the aneurysm sac on post-operative contrast-enhanced computed tomography scans (CT) in patients following endovascular repair (EVR) of abdominal aortic aneurysms (AAA). The aim of this study was to correlate the incidence of endoleaks with the presence of patent lumbar (LA) and inferior mesenteric arteries (IMA) as seen on pre-operative angiography. DESIGN, MATERIALS AND METHODS Forty-seven patients were assessed pre-operatively by both CT and angiography by a blinded radiologist prior to EVR of AAA. The number and size of patent vessels was recorded and correlated with the incidence of LA or IMA endoleaks on follow-up CT. Patent lumbar vessels were scored: 1 = small, 2 = medium, 3 = large. RESULTS Five patients were noted to have patent IMA on pre-operative angiography but none developed an endoleak. In this series, five patients had an endoleak due to a patent LA. The median score for patients with no endoleak was 1 (0-9) and for those with a lumbar endoleak 2 (0-5) (P = 0.26, Mann-Whitney U-test). The number of patent lumbar arteries was not predictive of a subsequent endoleak. Two out of nine (22 %) patients with large patent LA subsequently developed an endoleak. If a policy of pre-operative embolization on the basis of large patent LA had been adopted, seven patients would have had an unnecessary invasive procedure. CONCLUSION Pre-operative angiography to look for patent LA and IMAs is not required in patients undergoing EVR or AAA.

Comparison of morphologic features of intact and ruptured aneurysms of infrarenal abdominal aorta.

INTRODUCTION Endovascular aneurysm repair (EVAR) has been suggested as a technique to improve outcome of ruptured abdominal aortic aneurysm (AAA). Whether this technique becomes an established treatment will depend, in part, on the anatomy of ruptured AAA. METHODS The anatomy of intact and ruptured AAA seen in a university department of vascular surgery over 5 years was reviewed. Aneurysm anatomy was assessed with spiral computed tomographic angiography. Suitability for EVAR was assessed from the dimensions of the proximal neck and common iliac arteries. Neck length less than 15 mm, neck width greater than 30 mm, and common iliac artery diameter greater than 22 mm were declared unsuitable for EVAR. RESULTS Three hundred sixty-three patients with intact AAA and 46 with ruptured AAA were identified. Larger intact aneurysms were significantly associated with longer renal artery-bifurcation distance and more complex proximal neck architecture. In this sample, patients with ruptured AAA were more likely to have larger aneurysms with shorter and narrower proximal necks. Significantly more intact aneurysms were morphologically suitable for endovascular repair compared with ruptured AAA (78% vs 43%; P <.001). CONCLUSIONS Ruptured AAA are less likely to be suitable for endovascular repair than are intact AAA, most probably because of larger diameter at presentation. Open repair will likely remain the treatment of choice in most patients with ruptured AAA, because of current morphologic constraints of endovascular repair.

Stent-graft migration after endovascular repair of abdominal aortic aneurysm.

Purpose: To report the incidence of graft migration in patients after endovascular repair of abdominal aortic aneurysms (AAA) and assess the significance of neck diameter changes in patients with and without suprarenal stent implantation. Methods: The medical records and imaging studies of 176 consecutive patients (175 men; median age 71 years, range 48–88) who had endovascular AAA repair with the Nottingham aortomonoiliac system were reviewed. The following parameters were recorded: preoperative neck diameter and length, presence of intraoperative and late graft migrations, time to onset of late migration, length of late migration, and neck diameter changes in patients with documented late graft migration. The patients were divided into 2 groups based on the placement of an endograft with or without suprarenal bare stent fixation. Median follow-up was 15 months (range 1–48). Results: There were 15 (8.5%) graft migrations (6 intraoperative and 9 late). Of those, 14 (10.9%) were in the 128-patient infrarenal fixation group and 1 (2.1%) in the 48-patient suprarenal stent group. Median neck diameters on preoperative and postoperative computed tomography scans in patients with late migration were 22.2 mm and 23.0 mm, respectively (p>0.05). The median time to graft migration was 14 months after the original operation (range 6–36). Conclusions: Distal device migration occurred frequently with the Nottingham system. Late graft migration was not associated with neck enlargement. Endografts with a suprarenal stent may have a decreased incidence of graft migration.

Endovascular AAA Repair: Classification of Aneurysm Sac Volumetric Change Using Spiral Computed Tomographic Angiography

PURPOSE To classify and analyze the volumetric changes seen on spiral computed tomographic angiography (CTA) following endovascular abdominal aortic aneurysm (AAA) repair. METHODS Fifty patients (46 men; mean age 71 years, range 51-83) with >1 year of imaging follow-up were retrospectively selected. The volume of the aneurysm sac was calculated on standard CT workstations to obtain plots of volume changes over time. For the purpose of this study, a 10% change in sac volume was considered significant. RESULTS Over a mean 32-month follow-up, 256 CTA scans were performed; initial mean sac volume was 259 mL and initial mean AAA diameter was 6.5 cm. Six distinct patterns of volume change were recognized: group Ia (28 patients, 56%): progressive reduction in aneurysm sac volume; group Ib (3 patients, 6%): transient initial increase then same as Ia; group II (4 patients, 8%): no significant change; group IIIa (5 patients, 10%): late increase in volume; group IIIb (8 patients, 16%): progressive increase in volume; and group IV (2 patients, 4%): late reduction in volume after secondary intervention. Group III changes were associated with endoleak types I and III (p<0.0001). CONCLUSIONS This classification system of spiral CTA volumetric changes features 6 patterns with recognized clinical significance and predictive value for endoleaks. Group I is the ideal outcome when the aneurysm sac shrinks and often completely disappears, while group III is associated with types I and type III endoleak and should prompt further investigation. Long-term volumetric analysis of all patients is advised.

Intrarenal color duplex examination of aortic endograft patients with suprarenal stents.

PURPOSE To report an experience using intrarenal color duplex ultrasonography (ICDU) to detect high-grade renal artery stenosis in patients who had endovascular repair of abdominal aortic aneurysm (AAA) with suprarenal stent fixation. METHODS Twenty-eight patients (25 men; mean age 71 years, range 58-83) who had endovascular AAA repair with suprarenal stenting at least 3 months prior to commencement of this study were screened with ICDU. Acceleration time (AT), peak systolic velocity (PSV), end diastolic velocity (EDV), and resistive index (RI) were measured. The Doppler waveform was quantitatively scored on a scale from 0 to 4. AT >0.07 seconds, RI <0.45, or a Doppler waveform score of 0 or 1 (indicating loss of early systolic peak) were indicative of high-grade renal artery stenosis. RESULTS Median follow-up was 15.5 months (range 3-34). ICDU was successful in 54 (98%) of 55 kidneys scanned. No AT values exceeded 0.07 seconds, all RIs were >0.45, and no waveforms had loss of early systolic peak, indicating that no patient had evidence of high-grade renal artery stenosis. CONCLUSIONS ICDU is a simple and affordable method that seems well suited to periodic screening in patients with suprarenal stents. Longer follow-up with a larger number of patients is needed before definite conclusions can be drawn about the effect of suprarenal stenting on renal circulation.

Ex-Effects: An Empirical Reassessment of the Clientele Effect Using UK Data

This paper deals with two empirical aspects of the Elton and Gruber tax-induced clientele hypothesis. The first is the extent to which estimates of the central location of the ‘ex-dividend’ statistic, or marginal capitalisation of the dividend due, is influenced by different methods of estimation. The second is the degree to which the widely reported ‘dividend yield’ effect is a robust feature of the data, or whether its origins lie in aggregation effects or thin trading. The results suggest that the yield effect is much weaker than is generally claimed, bringing into question the simple tax-induced clientele hypothesis as the main explanation of cum-div to ex-div market value transitions.

Ex-Effects: Ex-Dividend-Ex—Rights Corroboration and the Implications for Valuation

Abstract The absurd prescriptive implications of ‘corner point’ solutions to financing and dividend theories, based on conventional models of tax perturbation, have led to an explosion of models based on information asymmetry or agency theory. A third line of enquiry, which posits that due to fundamental effects the nature of market valuation itself may be altered by taxation and other frictions, has received less attention. This paper addresses the valuation issue from an empirical perspective. The point at issue is whether the incidence of taxation is passed through market prices, as the Chicago ‘tight prior equilibrium’ view would seem to imply (the ‘conventional view’, on which much of the theory of finance is based); or whether market prices in some way reflect the tax positions of the investing clientele (the ‘capitalisation view’). The ‘window’ on the valuation process that is investigated empirically concerns the relationships between marginal market valuation changes for different ex-events, at t...

Color duplex ultrasonography of the superior mesenteric artery after placement of endografts with suprarenal stents.

After endovascular repair of abdominal aortic aneurysm with endografts with suprarenal stents, the proximal uncovered stent may cross the origin of the superior mesenteric artery. Effects on splanchnic circulation are unknown and may include development of stenosis at the vicinity of the stent. The criteria of high-grade superior mesenteric artery stenosis using color duplex ultrasonography have been previously reported. The purpose of this study is to examine the incidence of high-grade superior mesenteric artery stenosis in patients with endografts with suprarenal stents using color duplex ultrasonography. Candidates for the study were patients who had placement of an aortic endograft with a suprarenal stent and were able to undergo ultrasonography of the superior mesenteric artery. After reviewing computed tomography scans, patients who had the origin of the superior mesenteric artery crossed by the suprarenal stent underwent color duplex ultrasonography of this vessel. Presence of turbulence or narrowing of the superior mesenteric artery, or a peak systolic velocity greater than 2.75 m/sec, or an end-diastolic velocity greater than 0.45 m/sec were considered significant for the presence of high-grade superior mesenteric artery stenosis. There were 24 patients (21 males, three females), median age 71 years (range, 59-83). The suprarenal stent was crossing the superior mesenteric artery in 17 of 24 patients (71%). Color duplex ultrasound was technically successful in 13 of 17 (76%). The test was performed after a median follow-up of 9 months (range, 3 days to 34 months). No patient had evidence of turbulence or narrowing of the superior mesenteric artery during ultrasonography. The median peak systolic velocity was 0.92 m/sec (range, 0.53-1.21 m/sec). No patient had peak systolic velocity greater than 2.75 m/sec. The median end-diastolic velocity was 0.10 m/sec (range, 0.09-0.14 m/sec). No patient had end-diastolic velocity greater than 0.45 m/sec. Color duplex ultrasonography did not demonstrate the presence of high-grade superior mesenteric artery stenosis during early follow-up of patients with endografts with suprarenal stents. Longer follow-up of larger series of patients is needed to determine the long-term effects of suprarenal stents on splanchnic circulation.