CLINICAL STUDY
Impact on Pulmonary Arterial Pressures after Repeated Endovascular Thrombectomy of Dialysis Grafts: A Prospective Follow-up Study Mu-Yang Hsieh, MD, Chao-Lun Lai, MD, PhD, Yen-Wen Wu, MD, PhD, Lin Lin, MD, Miao-Chun Ho, RN, and Chih-Cheng Wu, MD
ABSTRACT Purpose: Vascular access thrombosis is a common complication of arteriovenous dialysis grafts that results in silent pulmonary embolism (PE) in a substantial proportion of patients. However, the impact of repeated PE on the pulmonary vasculature remains unclear. Materials and Methods: From January 2010 to April 2012, 110 patients undergoing maintenance hemodialysis via arteriovenous grafts were recruited. Hemodynamic assessments, including transthoracic echocardiography and right heart catheterization, were performed at baseline and after 1 year to evaluate the changes in pulmonary artery (PA) pressures and heart function. Results: Fifty-two patients completed the follow-up hemodynamic assessment at a median duration of 535 days and had at least one endovascular thrombectomy procedure (median of seven). There was no significant difference in mean PA pressures between baseline and the end of follow-up (23.1 mm Hg ⫾ 6.8 vs 21.6 mm Hg ⫾ 6.1; P ¼ .16). The change in mean PA pressure did not correlate with the number of thrombectomy procedures in the overall cohort (r ¼ 0.02, P ¼ .89) or in the subgroup with cardiopulmonary disease (r ¼ 0.30, P ¼ .14). The changes of mean PA pressure were not associated with number of thrombectomy procedures (β ¼ 0.03, P ¼ .89). Conclusions: Repeated endovascular thrombectomy procedures are not related to changes in PA pressure in the short term. The present results support the safety of endovascular thrombectomy in the pulmonary vasculature.
ABBREVIATIONS ESRD = end-stage renal disease, LVEF = left ventricular ejection fraction, PA = pulmonary artery, PE = pulmonary embolism
Thrombosis is a frequent complication of hemodialysis arteriovenous grafts in patients with end-stage renal disease (ESRD) (1). Thrombectomy of arteriovenous grafts is performed by endovascular or surgical methods. Endovascular methods are convenient and From the Cardiology Division (M.Y.H., C.L.L., L.L., C.C.W.) and Nursing Department (M.C.H.), National Taiwan University Hospital Hsinchu Branch, No. 25, Lane 442, Sec. 1, Jingguo Rd., Hsinchu City 300, Taiwan; College of Medicine (C.L.L., C.C.W.), National Taiwan University; School of Medicine (Y.W.W., C.C.W.), National Yang-Ming University; and Nuclear Medicine and Cardiovascular Center (Y.W.W.), Far Eastern Memorial Hospital, Taipei, Taiwan. Received March 18, 2014; final revision received July 14, 2014; accepted July 15, 2014. Address correspondence to C.C.W.; E-mail: [email protected] This study was supported by National Taiwan University Hospital, Hsinchu Branch, Grants DOH-99-HO-2022, 100-026-F, HCH101-12, and HCH102-33. None of the authors have identified a conflict of interest. & SIR, 2014 J Vasc Interv Radiol 2014; 25:1883–1889 http://dx.doi.org/10.1016/j.jvir.2014.07.018
effective in restoring dialysis access flow. According to the Dialysis Outcomes Quality Initiative Guidelines published in 1997 (2), endovascular methods have gradually replaced surgical methods as the therapy of choice for arteriovenous graft thrombosis (2,3). During endovascular thrombectomy, thrombus has usually been dissolved, fragmented, macerated, or aspirated within the vessel by a variety of techniques rather than being removed intact from the vessel (3,4). Iatrogenic pulmonary emboli may be unavoidable during endovascular thrombectomy, and perfusion defects in pulmonary vasculature after thrombectomy procedures has been demonstrated (5–9). Although the volume of thrombus is small, with rare incidences of acute symptomatic pulmonary embolism (PE), the cumulative effect of repeated iatrogenic emboli remains unknown (7,8,10,11). Patients undergoing hemodialysis have a high prevalence of pulmonary hypertension, and a variety of contributing factors have been proposed (12). Repeated
1884
’
Safety of Repeated Thrombectomy of Dialysis Grafts
procedure-related PE has been suspected as a possible cause of pulmonary hypertension, but only one retrospective study, by Harp et al (13), has demonstrated a lack of correlation between them. However, this does not mean that pulmonary artery (PA) pressure is not worsened by repeated PE. Currently, endovascular methods have been widely used to salvage thrombosis of hemodialysis grafts, without any formal literature regarding their effect on the pulmonary vasculature (12). Accordingly, it was hypothesized that repeated endovascular thrombectomy procedures may be associated with a change in PA pressure. The present prospective study was designed to investigate the effect of repeated endovascular thrombectomy procedures on the pulmonary vasculature.
MATERIALS AND METHODS Study Design Patients with dysfunctional hemodialysis vascular access were referred to our institution for endovascular intervention based on one or more of the following findings: clinical signs suggesting graft dysfunction, reduction in flow rate, or increased venous pressure during dialysis. Patients were invited to participate in the study if they (i) had received maintenance hemodialysis for at least 6 months (ii) via synthetic graft with (iii) no clinical evidence of fluid overload (stable dry weight for 2 wk, no peripheral edema by physical examination, and no pulmonary edema by chest radiograph) (14). Patients were excluded if they had a history of synthetic graft thrombosis in the previous 2 weeks, a history of acute coronary syndrome, or decompensated heart failure in the previous 3 months. All study participants received a baseline evaluation including the following: (i) review of medical records, angiographic reports, and hemodialysis records; (ii) physical examination; (iii) laboratory investigations; (iv) chest radiography; (v) echocardiography; and (vi) right heart catheterization (data collected by three of the investigators). They were then followed at their respective hemodialysis centers and referred for repeat fistulography or intervention based on the same criteria applied at enrollment. Clinical follow-up (including mortality, hospitalization, and vascular access events) was performed by a coordinating study nurse at the study institution by review of medical records, telephone contact, or interview with the patients and their referring nephrologists on three occasions at 6-month intervals. During the second year of follow-up, these patients were invited to participate in a hemodynamic follow-up (including echocardiography and right heart catheterization) by telephone contract or at their next access intervention. Only patients who completed the baseline and
Hsieh et al
’
JVIR
hemodynamic follow-up examination were included in the final analysis. The study followed the Declaration of Helsinki (edition 6, revised 2000) and was approved by the institutional review board of our hospital. Each participant gave their written informed consent before participating in the study.
Study Participants From January 2010 to April 2012, a total of 110 patients received a baseline evaluation. All participants completed the clinical follow-up evaluation. Fourteen patients died before the second hemodynamic assessment. The causes of death were as follows: two patients died from a fatal myocardial infarction, one died from a cerebral infarction, seven died from sepsis from various sources, and four died of unknown causes. All mortalities occurred more than 30 days after diagnostic or interventional procedures. Initially, patients were invited to participate in the hemodynamic follow-up examination by telephone contact but were reticent to receive hemodynamic follow-up. As a consequence, patients were invited to participate in the hemodynamic follow-up in the second year at their interventions for vascular access. Fourteen patients died before the second hemodynamic assessment, and another 42 declined our invitation and did not receive interventions at our hospital in the second year. Fifty-two patients (average age, 68 y ⫾ 10) completed the second assessment at the end of second year. Their median hemodialysis duration at enrollment was 24 months (25%–75% interquartile range, 10–56 mo). Twenty-six (50%) had a history of heart or pulmonary disease, including 26 (50%) with ischemic heart disease, three (6%) with valvular heart disease, four (8%) with heart failure, and one (2%) with chronic obstructive lung disease. No patient had a history of obstructive sleep apnea or obesity (defined by body mass index 4 30 kg/ m2). The baseline characteristics of patients who died and those who received or refused the second hemodynamic assessment are provided in Table 1.
Endovascular Thrombectomy Procedures All procedures were performed by one of seven interventional radiologists with 5–12 years of experience. A short sheath (Terumo, Tokyo, Japan) was placed at the graft near the venous junction, directed toward the arterial anastomosis; another short sheath was placed at the graft near the arterial junction, directed toward the venous anastomosis. After successfully traversing the occluded segment and crossing the arterial anastomosis, the arterial plug was dislodged by forceful withdrawal of the low-pressure inflated balloon catheter. Thrombi dislodged within the access segment were repeatedly aspirated via the sheath as much as possible. Mechanical thrombectomy devices (Arrow-Trerotola percutaneous
Volume 25
’
Number 12
’
December
’
2014
1885
Table 1 . Baseline Characteristics of Study Participants Based on Second Hemodynamic Assessment Characteristic
Received (n ¼ 52)
Refused (n ¼ 44)
Died (n ¼ 14)
68 ⫾ 10
67 ⫾ 12
69 ⫾ 14
32/20
25/19
6/8
Demographic Mean age (y) ⫾ SD Sex (F/M) HD duration (mo) Median
25
32
14
10–56 7 (13)
19–66 3 (7)
6–24 4 (29)
Hypertension Diabetes mellitus
34 (65) 27 (52)
20 (45) 23 (52)
10 (71) 9 (64)
Dyslipidemia
17 (33)
10 (23)
3 (21)
Ischemic heart disease Valvular heart disease
26 (50) 3 (6)
12 (27)* 6 (14)
3 (21)* 3 (21)
Heart failure
4 (8)
5 (11)
3 (21)
COPD Connective tissue disease
1 (2) 0
4 (9) 1 (2)
0 1 (7)
Pulmonary hypertension
26 (50)
21 (47)
5 (35) 1 (7)
IQR Current smoker Comorbid conditions
Medication Antiplatelet
16 (31)
7 (16)
Anticoagulant
2 (4)
1 (2)
0
ACEI/ARB β-Blocker
4 (8) 6 (12)
2 (5) 3 (7)
2 (14) 0
Calcium channel blocker
3 (6)
3 (7)
1 (7)
5 (10)
2 (5)
0
19 8–40
34 21–42
9 4–33
Nitrates Access factor Graft age (mo) Median IQR Right/left arm
9/43
12/32
1/13
Upper/lower arm Straight/loop graft
11/41 17/35
19/25* 20/22
1/13 2/12
Values in parentheses are percentages. ACEI ¼ angiotensin converting enzyme inhibitor, ARB ¼ angiotensin receptor blocker, COPD ¼ chronic obstructive pulmonary disease, HD ¼ hemodialysis, IQR ¼ interquartile range, PA ¼ pulmonary artery, SD ¼ standard deviation. *P o .05 vs patients who received second hemodynamic assessment.
thrombectomy device; Arrow, Reading, Pennsylvania) were used for large thrombus burden or wall-adherent thrombus that compromised blood flow. No additional pharmacologic lytic therapy was used. Balloon angioplasty of the venous anastomosis was performed after thrombectomy procedures. Diagnostic angiography was performed after restoration of flow, and the underlying stenosis was treated with balloon angioplasty. At our institution, stents or stent-grafts were used only for vessel rupture during the study period. Intravenous heparin (5,000 U) was administered during the procedure, and antiplatelet agents were prescribed for 3 days after the procedure.
Echocardiography and Right Heart Catheterization Transthoracic echocardiography was performed by two cardiologists who had at least 5 years of experience.
These investigations were carried out on a midweek nondialysis day. An ultrasound machine equipped with a 2.5-MHz probe (IE33; Philips, Andover, Massachusetts) was used. Recordings were made with the patients in a left lateral position during quiet respiration with the use of standard windows. Echo images of still frames and loops were stored in optical discs for offline analysis. Complete two-dimensional, M-mode, and Doppler echocardiographic studies were obtained from each patient and measured according to the American Society of Echocardiography guidelines (15,16). The average value of three consecutive cardiac cycles was obtained. Following transthoracic echocardiography, study participants underwent a simplified right heart catheterization procedure for the measurement of PA pressure. For the right heart catheterization, a 6-F, 5-cm sheath (Terumo) was placed in the venous limb of the hemodialysis graft and directed toward the outflow vein.
1886
’
Safety of Repeated Thrombectomy of Dialysis Grafts
A 5-F, 110-cm-long pigtail catheter (Terumo), coupled with a 0.035-inch guide wire (Terumo), was inserted via the sheath into the vascular access. Under fluoroscopic guidance, the catheter was advanced along the graft, outflow vein, central vein, and right heart, and finally into the main PA. After the pigtail catheter was placed in the main PA, pressure readings were recorded during pullback from the main PA, right ventricle, and right atrium. After the right heart and echocardiographic evaluation, the results were provided to the hemodialysis centers, and further evaluation was left to the discretion of the attending nephrologist.
Hsieh et al
’
JVIR
baseline and after follow-up. The McNemar test was used to compare categoric data (aortic or mitral regurgitation of at least moderate severity) at baseline and after follow-up. Multiple linear regression analysis was used to evaluate independent predictors of mean PA pressure change, including age, sex, comorbidities, shunt characteristics, and thrombectomy times. All statistical tests were two-tailed, and P values lower than .05 were considered statistically significant. All statistical analyses were performed by three of the investigators by using SPSS software (version 17.0; SPSS, Chicago, Illinois).
Definitions Pulmonary hypertension was defined as a mean PA pressure of at least 25 mm Hg. Ischemic heart disease was defined by a history of myocardial infarction, percutaneous coronary intervention, coronary artery bypass surgery, old myocardial infarction by electrocardiography, regional wall motion abnormality by echocardiography, or positive coronary angiography (ie, 4 50% stenosis in at least one major coronary artery), exercise stress test, or thallium stress test findings. Left ventricular dysfunction was defined as a left ventricular ejection fraction (LVEF) of less than 45%. Valvular heart disease was defined as moderate to severe stenosis and/or regurgitation of the mitral and/or aortic valve, according to the recommendation from the American Society of Echocardiography (16). Acute coronary syndrome was defined as new-onset cardiac symptoms with positive ischemic findings by 12-lead electrocardiography, or abnormally elevated cardiac troponin levels (17). Decompensated heart failure included patients with documented worsening of their LVEF by 5% or more, with clinical evidence of peripheral edema by physical examination or lung edema on chest radiography. Chronic obstructive lung disease was defined by a history of smoking, exertional dyspnea, and pulmonary function testing (forced expiratory volume in 1 s/forced vital capacity o 0.7) (18). Frequency of thrombectomy was defined as the total number of thrombectomy procedures divided by the duration of follow-up.
Statistical Analysis The main outcome variable was the change in mean PA pressure. Continuous variables in normal distribution are expressed as means ⫾ standard deviation, and variables not normally distributed are expressed as medians with 25%–75% interquartile range; categoric variables are expressed as percentages. Differences between groups were compared by using the independent t test for continuous variables and the χ2 test for categoric variables. Spearman correlation analysis was used to explain the relation between two continuous variables. Paired t tests were used to compare the echocardiographic and hemodynamic parameters at
RESULTS Fifty-two patients completed the hemodynamic followup assessment. The median follow-up duration was 535 days. All received at least one endovascular thrombectomy procedure (median of seven; interquartile range, 6–11). The clinical success rate was 98.1% and the major complication rate was 1.9% for thrombectomy procedures. Mechanical thrombectomy devices were used in 25 of 52 patients (48%). In the group of patients who received hemodynamic follow-up, no patient received surgical thrombectomy and only one patient underwent surgical graft removal as a result of graft infection. All patients who refused hemodynamic follow-up assessment received at least one endovascular procedure (median of five; interquartile range, three to nine). One patient in the refusal group received surgical revision as a result of frequent thrombosis; one patient who died had received permanent catheter insertion after graft thrombosis per the referring nephrologist. PA pressures in the 52 patients who received hemodynamic follow-up were successfully measured without complications by right heart catheterization at baseline and at follow-up. The baseline systolic, diastolic, and mean PA pressures were 36.6 mm Hg ⫾ 8.9, 12.9 mm Hg ⫾ 5.9, and 23.1 mm Hg ⫾ 6.8, respectively. Pulmonary hypertension was documented in 26 patients by right heart catheterization, and their systolic, diastolic, and mean PA pressures were 43.9 mm Hg ⫾ 7.7, 17.0 mm Hg ⫾ 4.8, and 28.8 mm Hg ⫾ 4.3, respectively. In the remaining 26 patients without pulmonary hypertension, the systolic, diastolic, and mean PA pressures were 30.2 mm Hg ⫾ 5.9, 9.2 mm Hg ⫾ 4.6, and 17.7 mm Hg ⫾ 4.4, respectively. In the echocardiography study, one patient had an LVEF less than 45%, two patients had moderate to severe mitral regurgitation, and one patient had moderate to severe aortic regurgitation. No patients had significant aortic stenosis, mitral stenosis, pulmonary valvular disease, or patent foramen ovale. Comparisons of echocardiographic and hemodynamic parameters at baseline and follow-up assessments are provided in Table 2. There was no difference in mean PA
Volume 25
’
Number 12
’
December
’
2014
1887
Table 2 . Echocardiographic and Hemodynamic Parameters at Baseline and Follow-up Assessments P Parameter
Baseline
Follow-up Value
Echocardiography LVIDD (mm)
44.4 ⫾ 6.2 44.8 ⫾ 5.2
.79
LVIDS (mm) IVS (mm)
29.1 ⫾ 5.2 27.3 ⫾ 4.7 11.0 ⫾ 2.4 13.2 ⫾ 1.9
.38 .03
PW (mm)
10.5 ⫾ 1.8 11.9 ⫾ 2.0
.05
LVEF (%) LV mass (g)
63.9 ⫾ 9.6 61.0 ⫾ 11.2 .48 234 ⫾ 63 283 ⫾ 70 .01
Moderate or severe MR
2 (3.8)
1 (0.9)
.99
Moderate or severe AR Right heart catheterization
1 (1.9)
0
.99
Systolic PA pressure (mm Hg) 36.6 ⫾ 8.9 35.9 ⫾ 9.5
.54
Diastolic PA pressure (mm Hg) 12.9 ⫾ 5.9 13.1 ⫾ 5.2 Mean PA pressure (mm Hg) 23.1 ⫾ 6.8 21.6 ⫾ 6.1
.74 .16
Values presented as means ⫾ standard deviation where applicable. Values in parentheses are percentages. AR = aortic regurgitation, IVS ¼ interventricular septum thickness, LVEF ¼ left ventricular ejection fraction, LVIDD ¼ left ventricular internal diameter during diastole, LVIDS ¼ left ventricular internal diameter during systole, MR = mitral regurgitation, PA ¼ pulmonary artery, PW ¼ posterior wall thickness.
pressures between baseline and follow-up studies (23.1 mm Hg ⫾ 6.8 vs 21.6 mm Hg ⫾ 6.1, respectively; P ¼ .16; Fig 1). When the change in mean PA pressure was plotted against the frequency of thrombectomy procedures, no correlation was identified between these two factors (r ¼ 0.02, P ¼ .89; Fig 2a). In the subgroup of patients with cardiac or pulmonary disease (defined as heart failure, ischemic heart disease, valvular heart disease, or chronic obstructive lung disease), the change in mean PA pressure did not correlate with the number of thrombectomy procedures (r ¼ 0.30, P ¼ .14; Fig 2b). Multivariate linear regression
Figure 1. Changes in mean pulmonary artery pressure at baseline and follow-up assessment. PAP ¼ pulmonary artery pressure.
revealed that changes in mean PA pressure were associated with heart failure (β ¼ 0.29, P ¼ .04) and chronic obstructive lung disease (β ¼ 0.43, P ¼ .03), but not with the number of thrombectomy procedures (β ¼ 0.19, P ¼ .39; Table 3). To address the concern that an a priori sample size calculation was not performed, post-hoc power calculations on the differences in mean PA pressures were performed. The standard deviations of mean PA pressures were 6.8 mm Hg at baseline and 6.1 mm Hg at follow-up. Assuming a correlation coefficient between baseline and follow-up of 0.8, the standard deviation of difference was estimated as 4.1 mm Hg. Assuming a difference of mean PA pressure of 2.0 mm Hg, the α-level would be 0.05 and power would be 90%. Therefore, a sample size of 47 participants would have 90% power to detect a difference in mean PA pressure of 2.0 mm Hg with a paired t test with a 0.05 two-sided significance level.
DISCUSSION Patients with ESRD, especially those with prosthetic grafts, are prone to repeated vascular access thrombosis. During the thrombectomy procedure, varying amounts of thrombus may be released into the pulmonary circulation. A number of scintigraphic studies have documented the presence of silent PE after thrombectomy procedures (7– 9,19). Although these subclinical emboli have been thought to have little or no effect on pulmonary circulation, their cumulative effect remains unknown. As many as 15%–20% of patients with an acute PE will have incomplete resolution of scintigraphic perfusion defects after the embolic event (20–22). Currently, there are no data demonstrating the hemodynamic effect of repeated iatrogenic PEs. In the present prospective analysis of PA pressure measured by right heart catheterization, we demonstrated that repeated endovascular thrombectomy is not related to changes in PA pressure. Given the widespread use of endovascular thrombectomy and high prevalence of pulmonary hypertension in patients with ESRD, our results provide evidence supporting the shortterm safety of repeated thrombectomy. The absence of a hemodynamic impact of repeated thrombectomy procedures shown by the present study may be explained as follows. Thrombus volume from synthetic grafts is usually much smaller than those from deep vein thrombosis, based on surgical pathology specimens (10,23). In addition, mechanical fragmentation and aspiration of thrombus is usually attempted before the venous stenosis is dilated. Therefore, the volume of residual thrombus is minimized before migrating into the pulmonary vascular bed, with a more favorable volume–to–surface area ratio for intrinsic fibrinolysis. Finally, thrombosis of hemodialysis grafts is usually discovered early, and the age and character of thrombus may be different from that of deep vein thrombosis (10). In consequence, these iatrogenic pulmonary emboli may
1888
’
Safety of Repeated Thrombectomy of Dialysis Grafts
Hsieh et al
’
JVIR
Figure 2. Scatterplots of the change in mean pulmonary artery pressure against the number of thrombectomy procedures from baseline to the end of follow-up. (a) All followed patients and (b) patients with cardiopulmonary disease. PAP ¼ pulmonary artery pressure.
Table 3 . Linear Regression Analysis of Factors Related to Pulmonary Artery Pressure Change Factor
β
Lower Bound
Upper Bound
P Value
Univariate Thrombectomy (frequency)
0.03
0.37
0.32
.89
0.29 0.43
0.33 0.01
15.15 0.19
.04 .03
0.19
0.64
0.27
.39
Multivariate Heart failure COPD Thrombectomy (frequency)
COPD ¼ chronic obstructive pulmonary disease.
undergo thrombolysis by active intrinsic fibrinolysis within the pulmonary vasculature before structural change in the pulmonary arteries develops. Chronic thromboembolic pulmonary hypertension may occur in patients after acute pulmonary embolic events (20,24). Although the incidence of chronic thromboembolic pulmonary hypertension is relatively low, distinct medical conditions in patients undergoing hemodialysis may predispose to the development of chronic thromboembolic pulmonary hypertension. In a retrospective case-control study by Harp et al (13), patients who had previously undergone percutaneous thrombectomy of thrombosed hemodialysis grafts did not have a higher prevalence of pulmonary hypertension compared with control subjects with ESRD. In contrast to such studies, which used only echocardiography to estimate PA pressure, we evaluated the clinical and hemodynamic impact of thrombectomy procedures by using transthoracic echocardiography and right heart catheterization, the latter representing the goldstandard method for diagnosis of pulmonary hypertension. Although the Doppler velocity of tricuspid regurgitation correlates well with PA pressure, a large regurgitation volume is needed for accurate measurement. In addition, noninvasive estimation of right atrial pressure is challenging and often inaccurate. In consequence, noninvasive measurement of PA pressure is not always feasible, and
over- or underestimation of PA pressure may occur in as many as 40% of individuals (25,26), which was not reliable or precise enough for our study purposes. Compared with the aforementioned retrospective case-control study (13), the prospective nature of the present study allowed a more precise calculation of the number of thrombectomy procedures. Our prospective design also allowed the analysis of cofactors that contributed to PA pressure change. The present study has several limitations. Only approximately half of the study participants received follow-up hemodynamic assessment. They had comparable characteristics as the study participants who declined follow-up hemodynamic assessment, except regarding a higher frequency of thrombectomy procedures. Although they may have been biased as a subgroup of patients prone to thrombosis, no difference in the change of PA pressure was demonstrated. In addition, the percentage of patients with preexisting heart or pulmonary disease was even smaller, and the safety in these vulnerable patients requires further clarification. Also, despite direct PA pressure measurement by right heart catheterization, variations in PA pressure over time may arise as a result of differences in fluid status or dry weight. Because of the small number of new events, the change in PA pressure was used as a surrogate marker to evaluate the effect of thrombectomy, instead of pulmonary hypertension. Finally, remote thrombectomy
Volume 25
’
Number 12
’
December
’
2014
procedures preceding enrollment could not be assessed precisely, which might have impact on the baseline and follow-up PA measurements. There was no change in PA pressure over an average follow-up of 18 months in the cohort of patients receiving hemodialysis and endovascular thrombectomy procedures. In addition, the number and frequency of thrombectomy procedures was not related to changes in PA pressure. Our results support the short-term safety of repeated endovascular thrombectomy in the pulmonary vasculature.
REFERENCES 1. Clinical practice guidelines for vascular access. Am J Kidney Dis 2006; 48(suppl 1):S176–S247. 2. NKF-DOQI clinical practice guidelines for vascular access. National Kidney Foundation–Dialysis Outcomes Quality Initiative. Am J Kidney Dis 1997; 30(suppl):S150–S191. 3. Bittl JA. Catheter interventions for hemodialysis fistulas and grafts. JACC Cardiovasc Interv 2010; 3:1–11. 4. Coentrao L, Bizarro P, Ribeiro C, Neto R, Pestana M. Percutaneous treatment of thrombosed arteriovenous fistulas: clinical and economic implications. Clin J Am Soc Nephrol 2010; 5:2245–2250. 5. Dolmatch BL, Gray RJ, Horton KM. Will iatrogenic pulmonary embolization be our pulmonary embarrassment? Radiology 1994; 191:615–617. 6. Weng FL, Berns JS. Complications of percutaneous treatment of thrombosed hemodialysis access grafts. Semin Dial 2003; 16:257–262. 7. Smits HF, Van Rijk PP, Van Isselt JW, Mali WP, Koomans HA, Blankestijn PJ. Pulmonary embolism after thrombolysis of hemodialysis grafts. J Am Soc Nephrol 1997; 8:1458–1461. 8. Swan TL, Smyth SH, Ruffenach SJ, Berman SS, Pond GD. Pulmonary embolism following hemodialysis access thrombolysis/thrombectomy. J Vasc Interv Radiol 1995; 6:683–686. 9. Kinney TB, Valji K, Rose SC, et al. Pulmonary embolism from pulsespray pharmacomechanical thrombolysis of clotted hemodialysis grafts: urokinase versus heparinized saline. J Vasc Interv Radiol 2000; 11: 1143–1152. 10. Winkler TA, Trerotola SO, Davidson DD, Milgrom ML. Study of thrombus from thrombosed hemodialysis access grafts. Radiology 1995; 197:461–465. 11. Trerotola SO, Johnson MS, Schauwecker DS, et al. Pulmonary emboli from pulse-spray and mechanical thrombolysis: evaluation with an animal dialysis-graft model. Radiology 1996; 200:169–176. 12. Zoccali C. Pulmonary hypertension in dialysis patients: a prevalent, risky but still uncharacterized disorder. Nephrol Dial Transplant 2012; 27: 3674–3677.
1889
13. Harp RJ, Stavropoulos SW, Wasserstein AG, Clark TW. Pulmonary hypertension among end-stage renal failure patients following hemodialysis access thrombectomy. Cardiovasc Intervent Radiol 2005; 28:17–22. 14. Charra B, Laurent G, Chazot C, et al. Clinical assessment of dry weight. Nephrol Dial Transplant 1996; 11(suppl 2):16–19. 15. Quinones MA, Otto CM, Stoddard M, Waggoner A, Zoghbi WA. Recommendations for quantification of Doppler echocardiography: a report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. J Am Soc Echocardiogr 2002; 15:167–184. 16. Zoghbi WA, Enriquez-Sarano M, Foster E, et al. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003; 16:777–802. 17. Luepker RV, Apple FS, Christenson RH, et al. Case definitions for acute coronary heart disease in epidemiology and clinical research studies: a statement from the AHA Council on Epidemiology and Prevention; AHA Statistics Committee; World Heart Federation Council on Epidemiology and Prevention; the European Society of Cardiology Working Group on Epidemiology and Prevention; Centers for Disease Control and Prevention; and the National Heart, Lung, and Blood Institute. Circulation 2003; 108:2543–2549. 18. Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS, GOLD Scientific Committee. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001; 163:1256–1276. 19. Beathard GA, Welch BR, Maidment HJ. Mechanical thrombolysis for the treatment of thrombosed hemodialysis access grafts. Radiology 1996; 200:711–716. 20. Becattini C, Agnelli G, Pesavento R, et al. Incidence of chronic thromboembolic pulmonary hypertension after a first episode of pulmonary embolism. Chest 2006; 130:172–175. 21. Wartski M, Collignon MA. Incomplete recovery of lung perfusion after 3 months in patients with acute pulmonary embolism treated with antithrombotic agents. J Nucl Med 2000; 41:1043–1048. 22. Pengo V, Lensing AWA, Prins MH, et al. Incidence of chronic thromboembolic pulmonary hypertension after pulmonary embolism. N Engl J Med 2004; 350:2257–2264. 23. Trerotola SO, Lund GB, Scheel PJ Jr, Savader SJ, Venbrux AC, Osterman FA Jr. Thrombosed dialysis access grafts: percutaneous mechanical declotting without urokinase. Radiology 1994; 191:721–726. 24. Ribeiro A, Lindmarker P, Johnsson H, Juhlin-Dannfelt A, Jorfeldt L. Pulmonary embolism: a follow-up study of the relation between the degree of right ventricle overload and the extent of perfusion defects. J Intern Med 1999; 245:601–610. 25. Fisher MR, Forfia PR, Chamera E, et al. Accuracy of Doppler echocardiography in the hemodynamic assessment of pulmonary hypertension. Am J Respir Crit Care Med 2009; 179:615–621. 26. Litwin SE. Noninvasive assessment of pulmonary artery pressures: moving beyond tricuspid regurgitation velocities. Circ Cardiovasc Imaging 2010; 3:132–133.
Impact on Pulmonary Arterial Pressures after Repeated Endovascular Thrombectomy of Dialysis Grafts: A Prospective Follow-up Study Mu-Yang Hsieh, MD, Chao-Lun Lai, MD, PhD, Yen-Wen Wu, MD, PhD, Lin Lin, MD, Miao-Chun Ho, RN, and Chih-Cheng Wu, MD
ABSTRACT Purpose: Vascular access thrombosis is a common complication of arteriovenous dialysis grafts that results in silent pulmonary embolism (PE) in a substantial proportion of patients. However, the impact of repeated PE on the pulmonary vasculature remains unclear. Materials and Methods: From January 2010 to April 2012, 110 patients undergoing maintenance hemodialysis via arteriovenous grafts were recruited. Hemodynamic assessments, including transthoracic echocardiography and right heart catheterization, were performed at baseline and after 1 year to evaluate the changes in pulmonary artery (PA) pressures and heart function. Results: Fifty-two patients completed the follow-up hemodynamic assessment at a median duration of 535 days and had at least one endovascular thrombectomy procedure (median of seven). There was no significant difference in mean PA pressures between baseline and the end of follow-up (23.1 mm Hg ⫾ 6.8 vs 21.6 mm Hg ⫾ 6.1; P ¼ .16). The change in mean PA pressure did not correlate with the number of thrombectomy procedures in the overall cohort (r ¼ 0.02, P ¼ .89) or in the subgroup with cardiopulmonary disease (r ¼ 0.30, P ¼ .14). The changes of mean PA pressure were not associated with number of thrombectomy procedures (β ¼ 0.03, P ¼ .89). Conclusions: Repeated endovascular thrombectomy procedures are not related to changes in PA pressure in the short term. The present results support the safety of endovascular thrombectomy in the pulmonary vasculature.
ABBREVIATIONS ESRD = end-stage renal disease, LVEF = left ventricular ejection fraction, PA = pulmonary artery, PE = pulmonary embolism
Thrombosis is a frequent complication of hemodialysis arteriovenous grafts in patients with end-stage renal disease (ESRD) (1). Thrombectomy of arteriovenous grafts is performed by endovascular or surgical methods. Endovascular methods are convenient and From the Cardiology Division (M.Y.H., C.L.L., L.L., C.C.W.) and Nursing Department (M.C.H.), National Taiwan University Hospital Hsinchu Branch, No. 25, Lane 442, Sec. 1, Jingguo Rd., Hsinchu City 300, Taiwan; College of Medicine (C.L.L., C.C.W.), National Taiwan University; School of Medicine (Y.W.W., C.C.W.), National Yang-Ming University; and Nuclear Medicine and Cardiovascular Center (Y.W.W.), Far Eastern Memorial Hospital, Taipei, Taiwan. Received March 18, 2014; final revision received July 14, 2014; accepted July 15, 2014. Address correspondence to C.C.W.; E-mail: [email protected] This study was supported by National Taiwan University Hospital, Hsinchu Branch, Grants DOH-99-HO-2022, 100-026-F, HCH101-12, and HCH102-33. None of the authors have identified a conflict of interest. & SIR, 2014 J Vasc Interv Radiol 2014; 25:1883–1889 http://dx.doi.org/10.1016/j.jvir.2014.07.018
effective in restoring dialysis access flow. According to the Dialysis Outcomes Quality Initiative Guidelines published in 1997 (2), endovascular methods have gradually replaced surgical methods as the therapy of choice for arteriovenous graft thrombosis (2,3). During endovascular thrombectomy, thrombus has usually been dissolved, fragmented, macerated, or aspirated within the vessel by a variety of techniques rather than being removed intact from the vessel (3,4). Iatrogenic pulmonary emboli may be unavoidable during endovascular thrombectomy, and perfusion defects in pulmonary vasculature after thrombectomy procedures has been demonstrated (5–9). Although the volume of thrombus is small, with rare incidences of acute symptomatic pulmonary embolism (PE), the cumulative effect of repeated iatrogenic emboli remains unknown (7,8,10,11). Patients undergoing hemodialysis have a high prevalence of pulmonary hypertension, and a variety of contributing factors have been proposed (12). Repeated
1884
’
Safety of Repeated Thrombectomy of Dialysis Grafts
procedure-related PE has been suspected as a possible cause of pulmonary hypertension, but only one retrospective study, by Harp et al (13), has demonstrated a lack of correlation between them. However, this does not mean that pulmonary artery (PA) pressure is not worsened by repeated PE. Currently, endovascular methods have been widely used to salvage thrombosis of hemodialysis grafts, without any formal literature regarding their effect on the pulmonary vasculature (12). Accordingly, it was hypothesized that repeated endovascular thrombectomy procedures may be associated with a change in PA pressure. The present prospective study was designed to investigate the effect of repeated endovascular thrombectomy procedures on the pulmonary vasculature.
MATERIALS AND METHODS Study Design Patients with dysfunctional hemodialysis vascular access were referred to our institution for endovascular intervention based on one or more of the following findings: clinical signs suggesting graft dysfunction, reduction in flow rate, or increased venous pressure during dialysis. Patients were invited to participate in the study if they (i) had received maintenance hemodialysis for at least 6 months (ii) via synthetic graft with (iii) no clinical evidence of fluid overload (stable dry weight for 2 wk, no peripheral edema by physical examination, and no pulmonary edema by chest radiograph) (14). Patients were excluded if they had a history of synthetic graft thrombosis in the previous 2 weeks, a history of acute coronary syndrome, or decompensated heart failure in the previous 3 months. All study participants received a baseline evaluation including the following: (i) review of medical records, angiographic reports, and hemodialysis records; (ii) physical examination; (iii) laboratory investigations; (iv) chest radiography; (v) echocardiography; and (vi) right heart catheterization (data collected by three of the investigators). They were then followed at their respective hemodialysis centers and referred for repeat fistulography or intervention based on the same criteria applied at enrollment. Clinical follow-up (including mortality, hospitalization, and vascular access events) was performed by a coordinating study nurse at the study institution by review of medical records, telephone contact, or interview with the patients and their referring nephrologists on three occasions at 6-month intervals. During the second year of follow-up, these patients were invited to participate in a hemodynamic follow-up (including echocardiography and right heart catheterization) by telephone contract or at their next access intervention. Only patients who completed the baseline and
Hsieh et al
’
JVIR
hemodynamic follow-up examination were included in the final analysis. The study followed the Declaration of Helsinki (edition 6, revised 2000) and was approved by the institutional review board of our hospital. Each participant gave their written informed consent before participating in the study.
Study Participants From January 2010 to April 2012, a total of 110 patients received a baseline evaluation. All participants completed the clinical follow-up evaluation. Fourteen patients died before the second hemodynamic assessment. The causes of death were as follows: two patients died from a fatal myocardial infarction, one died from a cerebral infarction, seven died from sepsis from various sources, and four died of unknown causes. All mortalities occurred more than 30 days after diagnostic or interventional procedures. Initially, patients were invited to participate in the hemodynamic follow-up examination by telephone contact but were reticent to receive hemodynamic follow-up. As a consequence, patients were invited to participate in the hemodynamic follow-up in the second year at their interventions for vascular access. Fourteen patients died before the second hemodynamic assessment, and another 42 declined our invitation and did not receive interventions at our hospital in the second year. Fifty-two patients (average age, 68 y ⫾ 10) completed the second assessment at the end of second year. Their median hemodialysis duration at enrollment was 24 months (25%–75% interquartile range, 10–56 mo). Twenty-six (50%) had a history of heart or pulmonary disease, including 26 (50%) with ischemic heart disease, three (6%) with valvular heart disease, four (8%) with heart failure, and one (2%) with chronic obstructive lung disease. No patient had a history of obstructive sleep apnea or obesity (defined by body mass index 4 30 kg/ m2). The baseline characteristics of patients who died and those who received or refused the second hemodynamic assessment are provided in Table 1.
Endovascular Thrombectomy Procedures All procedures were performed by one of seven interventional radiologists with 5–12 years of experience. A short sheath (Terumo, Tokyo, Japan) was placed at the graft near the venous junction, directed toward the arterial anastomosis; another short sheath was placed at the graft near the arterial junction, directed toward the venous anastomosis. After successfully traversing the occluded segment and crossing the arterial anastomosis, the arterial plug was dislodged by forceful withdrawal of the low-pressure inflated balloon catheter. Thrombi dislodged within the access segment were repeatedly aspirated via the sheath as much as possible. Mechanical thrombectomy devices (Arrow-Trerotola percutaneous
Volume 25
’
Number 12
’
December
’
2014
1885
Table 1 . Baseline Characteristics of Study Participants Based on Second Hemodynamic Assessment Characteristic
Received (n ¼ 52)
Refused (n ¼ 44)
Died (n ¼ 14)
68 ⫾ 10
67 ⫾ 12
69 ⫾ 14
32/20
25/19
6/8
Demographic Mean age (y) ⫾ SD Sex (F/M) HD duration (mo) Median
25
32
14
10–56 7 (13)
19–66 3 (7)
6–24 4 (29)
Hypertension Diabetes mellitus
34 (65) 27 (52)
20 (45) 23 (52)
10 (71) 9 (64)
Dyslipidemia
17 (33)
10 (23)
3 (21)
Ischemic heart disease Valvular heart disease
26 (50) 3 (6)
12 (27)* 6 (14)
3 (21)* 3 (21)
Heart failure
4 (8)
5 (11)
3 (21)
COPD Connective tissue disease
1 (2) 0
4 (9) 1 (2)
0 1 (7)
Pulmonary hypertension
26 (50)
21 (47)
5 (35) 1 (7)
IQR Current smoker Comorbid conditions
Medication Antiplatelet
16 (31)
7 (16)
Anticoagulant
2 (4)
1 (2)
0
ACEI/ARB β-Blocker
4 (8) 6 (12)
2 (5) 3 (7)
2 (14) 0
Calcium channel blocker
3 (6)
3 (7)
1 (7)
5 (10)
2 (5)
0
19 8–40
34 21–42
9 4–33
Nitrates Access factor Graft age (mo) Median IQR Right/left arm
9/43
12/32
1/13
Upper/lower arm Straight/loop graft
11/41 17/35
19/25* 20/22
1/13 2/12
Values in parentheses are percentages. ACEI ¼ angiotensin converting enzyme inhibitor, ARB ¼ angiotensin receptor blocker, COPD ¼ chronic obstructive pulmonary disease, HD ¼ hemodialysis, IQR ¼ interquartile range, PA ¼ pulmonary artery, SD ¼ standard deviation. *P o .05 vs patients who received second hemodynamic assessment.
thrombectomy device; Arrow, Reading, Pennsylvania) were used for large thrombus burden or wall-adherent thrombus that compromised blood flow. No additional pharmacologic lytic therapy was used. Balloon angioplasty of the venous anastomosis was performed after thrombectomy procedures. Diagnostic angiography was performed after restoration of flow, and the underlying stenosis was treated with balloon angioplasty. At our institution, stents or stent-grafts were used only for vessel rupture during the study period. Intravenous heparin (5,000 U) was administered during the procedure, and antiplatelet agents were prescribed for 3 days after the procedure.
Echocardiography and Right Heart Catheterization Transthoracic echocardiography was performed by two cardiologists who had at least 5 years of experience.
These investigations were carried out on a midweek nondialysis day. An ultrasound machine equipped with a 2.5-MHz probe (IE33; Philips, Andover, Massachusetts) was used. Recordings were made with the patients in a left lateral position during quiet respiration with the use of standard windows. Echo images of still frames and loops were stored in optical discs for offline analysis. Complete two-dimensional, M-mode, and Doppler echocardiographic studies were obtained from each patient and measured according to the American Society of Echocardiography guidelines (15,16). The average value of three consecutive cardiac cycles was obtained. Following transthoracic echocardiography, study participants underwent a simplified right heart catheterization procedure for the measurement of PA pressure. For the right heart catheterization, a 6-F, 5-cm sheath (Terumo) was placed in the venous limb of the hemodialysis graft and directed toward the outflow vein.
1886
’
Safety of Repeated Thrombectomy of Dialysis Grafts
A 5-F, 110-cm-long pigtail catheter (Terumo), coupled with a 0.035-inch guide wire (Terumo), was inserted via the sheath into the vascular access. Under fluoroscopic guidance, the catheter was advanced along the graft, outflow vein, central vein, and right heart, and finally into the main PA. After the pigtail catheter was placed in the main PA, pressure readings were recorded during pullback from the main PA, right ventricle, and right atrium. After the right heart and echocardiographic evaluation, the results were provided to the hemodialysis centers, and further evaluation was left to the discretion of the attending nephrologist.
Hsieh et al
’
JVIR
baseline and after follow-up. The McNemar test was used to compare categoric data (aortic or mitral regurgitation of at least moderate severity) at baseline and after follow-up. Multiple linear regression analysis was used to evaluate independent predictors of mean PA pressure change, including age, sex, comorbidities, shunt characteristics, and thrombectomy times. All statistical tests were two-tailed, and P values lower than .05 were considered statistically significant. All statistical analyses were performed by three of the investigators by using SPSS software (version 17.0; SPSS, Chicago, Illinois).
Definitions Pulmonary hypertension was defined as a mean PA pressure of at least 25 mm Hg. Ischemic heart disease was defined by a history of myocardial infarction, percutaneous coronary intervention, coronary artery bypass surgery, old myocardial infarction by electrocardiography, regional wall motion abnormality by echocardiography, or positive coronary angiography (ie, 4 50% stenosis in at least one major coronary artery), exercise stress test, or thallium stress test findings. Left ventricular dysfunction was defined as a left ventricular ejection fraction (LVEF) of less than 45%. Valvular heart disease was defined as moderate to severe stenosis and/or regurgitation of the mitral and/or aortic valve, according to the recommendation from the American Society of Echocardiography (16). Acute coronary syndrome was defined as new-onset cardiac symptoms with positive ischemic findings by 12-lead electrocardiography, or abnormally elevated cardiac troponin levels (17). Decompensated heart failure included patients with documented worsening of their LVEF by 5% or more, with clinical evidence of peripheral edema by physical examination or lung edema on chest radiography. Chronic obstructive lung disease was defined by a history of smoking, exertional dyspnea, and pulmonary function testing (forced expiratory volume in 1 s/forced vital capacity o 0.7) (18). Frequency of thrombectomy was defined as the total number of thrombectomy procedures divided by the duration of follow-up.
Statistical Analysis The main outcome variable was the change in mean PA pressure. Continuous variables in normal distribution are expressed as means ⫾ standard deviation, and variables not normally distributed are expressed as medians with 25%–75% interquartile range; categoric variables are expressed as percentages. Differences between groups were compared by using the independent t test for continuous variables and the χ2 test for categoric variables. Spearman correlation analysis was used to explain the relation between two continuous variables. Paired t tests were used to compare the echocardiographic and hemodynamic parameters at
RESULTS Fifty-two patients completed the hemodynamic followup assessment. The median follow-up duration was 535 days. All received at least one endovascular thrombectomy procedure (median of seven; interquartile range, 6–11). The clinical success rate was 98.1% and the major complication rate was 1.9% for thrombectomy procedures. Mechanical thrombectomy devices were used in 25 of 52 patients (48%). In the group of patients who received hemodynamic follow-up, no patient received surgical thrombectomy and only one patient underwent surgical graft removal as a result of graft infection. All patients who refused hemodynamic follow-up assessment received at least one endovascular procedure (median of five; interquartile range, three to nine). One patient in the refusal group received surgical revision as a result of frequent thrombosis; one patient who died had received permanent catheter insertion after graft thrombosis per the referring nephrologist. PA pressures in the 52 patients who received hemodynamic follow-up were successfully measured without complications by right heart catheterization at baseline and at follow-up. The baseline systolic, diastolic, and mean PA pressures were 36.6 mm Hg ⫾ 8.9, 12.9 mm Hg ⫾ 5.9, and 23.1 mm Hg ⫾ 6.8, respectively. Pulmonary hypertension was documented in 26 patients by right heart catheterization, and their systolic, diastolic, and mean PA pressures were 43.9 mm Hg ⫾ 7.7, 17.0 mm Hg ⫾ 4.8, and 28.8 mm Hg ⫾ 4.3, respectively. In the remaining 26 patients without pulmonary hypertension, the systolic, diastolic, and mean PA pressures were 30.2 mm Hg ⫾ 5.9, 9.2 mm Hg ⫾ 4.6, and 17.7 mm Hg ⫾ 4.4, respectively. In the echocardiography study, one patient had an LVEF less than 45%, two patients had moderate to severe mitral regurgitation, and one patient had moderate to severe aortic regurgitation. No patients had significant aortic stenosis, mitral stenosis, pulmonary valvular disease, or patent foramen ovale. Comparisons of echocardiographic and hemodynamic parameters at baseline and follow-up assessments are provided in Table 2. There was no difference in mean PA
Volume 25
’
Number 12
’
December
’
2014
1887
Table 2 . Echocardiographic and Hemodynamic Parameters at Baseline and Follow-up Assessments P Parameter
Baseline
Follow-up Value
Echocardiography LVIDD (mm)
44.4 ⫾ 6.2 44.8 ⫾ 5.2
.79
LVIDS (mm) IVS (mm)
29.1 ⫾ 5.2 27.3 ⫾ 4.7 11.0 ⫾ 2.4 13.2 ⫾ 1.9
.38 .03
PW (mm)
10.5 ⫾ 1.8 11.9 ⫾ 2.0
.05
LVEF (%) LV mass (g)
63.9 ⫾ 9.6 61.0 ⫾ 11.2 .48 234 ⫾ 63 283 ⫾ 70 .01
Moderate or severe MR
2 (3.8)
1 (0.9)
.99
Moderate or severe AR Right heart catheterization
1 (1.9)
0
.99
Systolic PA pressure (mm Hg) 36.6 ⫾ 8.9 35.9 ⫾ 9.5
.54
Diastolic PA pressure (mm Hg) 12.9 ⫾ 5.9 13.1 ⫾ 5.2 Mean PA pressure (mm Hg) 23.1 ⫾ 6.8 21.6 ⫾ 6.1
.74 .16
Values presented as means ⫾ standard deviation where applicable. Values in parentheses are percentages. AR = aortic regurgitation, IVS ¼ interventricular septum thickness, LVEF ¼ left ventricular ejection fraction, LVIDD ¼ left ventricular internal diameter during diastole, LVIDS ¼ left ventricular internal diameter during systole, MR = mitral regurgitation, PA ¼ pulmonary artery, PW ¼ posterior wall thickness.
pressures between baseline and follow-up studies (23.1 mm Hg ⫾ 6.8 vs 21.6 mm Hg ⫾ 6.1, respectively; P ¼ .16; Fig 1). When the change in mean PA pressure was plotted against the frequency of thrombectomy procedures, no correlation was identified between these two factors (r ¼ 0.02, P ¼ .89; Fig 2a). In the subgroup of patients with cardiac or pulmonary disease (defined as heart failure, ischemic heart disease, valvular heart disease, or chronic obstructive lung disease), the change in mean PA pressure did not correlate with the number of thrombectomy procedures (r ¼ 0.30, P ¼ .14; Fig 2b). Multivariate linear regression
Figure 1. Changes in mean pulmonary artery pressure at baseline and follow-up assessment. PAP ¼ pulmonary artery pressure.
revealed that changes in mean PA pressure were associated with heart failure (β ¼ 0.29, P ¼ .04) and chronic obstructive lung disease (β ¼ 0.43, P ¼ .03), but not with the number of thrombectomy procedures (β ¼ 0.19, P ¼ .39; Table 3). To address the concern that an a priori sample size calculation was not performed, post-hoc power calculations on the differences in mean PA pressures were performed. The standard deviations of mean PA pressures were 6.8 mm Hg at baseline and 6.1 mm Hg at follow-up. Assuming a correlation coefficient between baseline and follow-up of 0.8, the standard deviation of difference was estimated as 4.1 mm Hg. Assuming a difference of mean PA pressure of 2.0 mm Hg, the α-level would be 0.05 and power would be 90%. Therefore, a sample size of 47 participants would have 90% power to detect a difference in mean PA pressure of 2.0 mm Hg with a paired t test with a 0.05 two-sided significance level.
DISCUSSION Patients with ESRD, especially those with prosthetic grafts, are prone to repeated vascular access thrombosis. During the thrombectomy procedure, varying amounts of thrombus may be released into the pulmonary circulation. A number of scintigraphic studies have documented the presence of silent PE after thrombectomy procedures (7– 9,19). Although these subclinical emboli have been thought to have little or no effect on pulmonary circulation, their cumulative effect remains unknown. As many as 15%–20% of patients with an acute PE will have incomplete resolution of scintigraphic perfusion defects after the embolic event (20–22). Currently, there are no data demonstrating the hemodynamic effect of repeated iatrogenic PEs. In the present prospective analysis of PA pressure measured by right heart catheterization, we demonstrated that repeated endovascular thrombectomy is not related to changes in PA pressure. Given the widespread use of endovascular thrombectomy and high prevalence of pulmonary hypertension in patients with ESRD, our results provide evidence supporting the shortterm safety of repeated thrombectomy. The absence of a hemodynamic impact of repeated thrombectomy procedures shown by the present study may be explained as follows. Thrombus volume from synthetic grafts is usually much smaller than those from deep vein thrombosis, based on surgical pathology specimens (10,23). In addition, mechanical fragmentation and aspiration of thrombus is usually attempted before the venous stenosis is dilated. Therefore, the volume of residual thrombus is minimized before migrating into the pulmonary vascular bed, with a more favorable volume–to–surface area ratio for intrinsic fibrinolysis. Finally, thrombosis of hemodialysis grafts is usually discovered early, and the age and character of thrombus may be different from that of deep vein thrombosis (10). In consequence, these iatrogenic pulmonary emboli may
1888
’
Safety of Repeated Thrombectomy of Dialysis Grafts
Hsieh et al
’
JVIR
Figure 2. Scatterplots of the change in mean pulmonary artery pressure against the number of thrombectomy procedures from baseline to the end of follow-up. (a) All followed patients and (b) patients with cardiopulmonary disease. PAP ¼ pulmonary artery pressure.
Table 3 . Linear Regression Analysis of Factors Related to Pulmonary Artery Pressure Change Factor
β
Lower Bound
Upper Bound
P Value
Univariate Thrombectomy (frequency)
0.03
0.37
0.32
.89
0.29 0.43
0.33 0.01
15.15 0.19
.04 .03
0.19
0.64
0.27
.39
Multivariate Heart failure COPD Thrombectomy (frequency)
COPD ¼ chronic obstructive pulmonary disease.
undergo thrombolysis by active intrinsic fibrinolysis within the pulmonary vasculature before structural change in the pulmonary arteries develops. Chronic thromboembolic pulmonary hypertension may occur in patients after acute pulmonary embolic events (20,24). Although the incidence of chronic thromboembolic pulmonary hypertension is relatively low, distinct medical conditions in patients undergoing hemodialysis may predispose to the development of chronic thromboembolic pulmonary hypertension. In a retrospective case-control study by Harp et al (13), patients who had previously undergone percutaneous thrombectomy of thrombosed hemodialysis grafts did not have a higher prevalence of pulmonary hypertension compared with control subjects with ESRD. In contrast to such studies, which used only echocardiography to estimate PA pressure, we evaluated the clinical and hemodynamic impact of thrombectomy procedures by using transthoracic echocardiography and right heart catheterization, the latter representing the goldstandard method for diagnosis of pulmonary hypertension. Although the Doppler velocity of tricuspid regurgitation correlates well with PA pressure, a large regurgitation volume is needed for accurate measurement. In addition, noninvasive estimation of right atrial pressure is challenging and often inaccurate. In consequence, noninvasive measurement of PA pressure is not always feasible, and
over- or underestimation of PA pressure may occur in as many as 40% of individuals (25,26), which was not reliable or precise enough for our study purposes. Compared with the aforementioned retrospective case-control study (13), the prospective nature of the present study allowed a more precise calculation of the number of thrombectomy procedures. Our prospective design also allowed the analysis of cofactors that contributed to PA pressure change. The present study has several limitations. Only approximately half of the study participants received follow-up hemodynamic assessment. They had comparable characteristics as the study participants who declined follow-up hemodynamic assessment, except regarding a higher frequency of thrombectomy procedures. Although they may have been biased as a subgroup of patients prone to thrombosis, no difference in the change of PA pressure was demonstrated. In addition, the percentage of patients with preexisting heart or pulmonary disease was even smaller, and the safety in these vulnerable patients requires further clarification. Also, despite direct PA pressure measurement by right heart catheterization, variations in PA pressure over time may arise as a result of differences in fluid status or dry weight. Because of the small number of new events, the change in PA pressure was used as a surrogate marker to evaluate the effect of thrombectomy, instead of pulmonary hypertension. Finally, remote thrombectomy
Volume 25
’
Number 12
’
December
’
2014
procedures preceding enrollment could not be assessed precisely, which might have impact on the baseline and follow-up PA measurements. There was no change in PA pressure over an average follow-up of 18 months in the cohort of patients receiving hemodialysis and endovascular thrombectomy procedures. In addition, the number and frequency of thrombectomy procedures was not related to changes in PA pressure. Our results support the short-term safety of repeated endovascular thrombectomy in the pulmonary vasculature.
REFERENCES 1. Clinical practice guidelines for vascular access. Am J Kidney Dis 2006; 48(suppl 1):S176–S247. 2. NKF-DOQI clinical practice guidelines for vascular access. National Kidney Foundation–Dialysis Outcomes Quality Initiative. Am J Kidney Dis 1997; 30(suppl):S150–S191. 3. Bittl JA. Catheter interventions for hemodialysis fistulas and grafts. JACC Cardiovasc Interv 2010; 3:1–11. 4. Coentrao L, Bizarro P, Ribeiro C, Neto R, Pestana M. Percutaneous treatment of thrombosed arteriovenous fistulas: clinical and economic implications. Clin J Am Soc Nephrol 2010; 5:2245–2250. 5. Dolmatch BL, Gray RJ, Horton KM. Will iatrogenic pulmonary embolization be our pulmonary embarrassment? Radiology 1994; 191:615–617. 6. Weng FL, Berns JS. Complications of percutaneous treatment of thrombosed hemodialysis access grafts. Semin Dial 2003; 16:257–262. 7. Smits HF, Van Rijk PP, Van Isselt JW, Mali WP, Koomans HA, Blankestijn PJ. Pulmonary embolism after thrombolysis of hemodialysis grafts. J Am Soc Nephrol 1997; 8:1458–1461. 8. Swan TL, Smyth SH, Ruffenach SJ, Berman SS, Pond GD. Pulmonary embolism following hemodialysis access thrombolysis/thrombectomy. J Vasc Interv Radiol 1995; 6:683–686. 9. Kinney TB, Valji K, Rose SC, et al. Pulmonary embolism from pulsespray pharmacomechanical thrombolysis of clotted hemodialysis grafts: urokinase versus heparinized saline. J Vasc Interv Radiol 2000; 11: 1143–1152. 10. Winkler TA, Trerotola SO, Davidson DD, Milgrom ML. Study of thrombus from thrombosed hemodialysis access grafts. Radiology 1995; 197:461–465. 11. Trerotola SO, Johnson MS, Schauwecker DS, et al. Pulmonary emboli from pulse-spray and mechanical thrombolysis: evaluation with an animal dialysis-graft model. Radiology 1996; 200:169–176. 12. Zoccali C. Pulmonary hypertension in dialysis patients: a prevalent, risky but still uncharacterized disorder. Nephrol Dial Transplant 2012; 27: 3674–3677.
1889
13. Harp RJ, Stavropoulos SW, Wasserstein AG, Clark TW. Pulmonary hypertension among end-stage renal failure patients following hemodialysis access thrombectomy. Cardiovasc Intervent Radiol 2005; 28:17–22. 14. Charra B, Laurent G, Chazot C, et al. Clinical assessment of dry weight. Nephrol Dial Transplant 1996; 11(suppl 2):16–19. 15. Quinones MA, Otto CM, Stoddard M, Waggoner A, Zoghbi WA. Recommendations for quantification of Doppler echocardiography: a report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. J Am Soc Echocardiogr 2002; 15:167–184. 16. Zoghbi WA, Enriquez-Sarano M, Foster E, et al. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003; 16:777–802. 17. Luepker RV, Apple FS, Christenson RH, et al. Case definitions for acute coronary heart disease in epidemiology and clinical research studies: a statement from the AHA Council on Epidemiology and Prevention; AHA Statistics Committee; World Heart Federation Council on Epidemiology and Prevention; the European Society of Cardiology Working Group on Epidemiology and Prevention; Centers for Disease Control and Prevention; and the National Heart, Lung, and Blood Institute. Circulation 2003; 108:2543–2549. 18. Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS, GOLD Scientific Committee. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001; 163:1256–1276. 19. Beathard GA, Welch BR, Maidment HJ. Mechanical thrombolysis for the treatment of thrombosed hemodialysis access grafts. Radiology 1996; 200:711–716. 20. Becattini C, Agnelli G, Pesavento R, et al. Incidence of chronic thromboembolic pulmonary hypertension after a first episode of pulmonary embolism. Chest 2006; 130:172–175. 21. Wartski M, Collignon MA. Incomplete recovery of lung perfusion after 3 months in patients with acute pulmonary embolism treated with antithrombotic agents. J Nucl Med 2000; 41:1043–1048. 22. Pengo V, Lensing AWA, Prins MH, et al. Incidence of chronic thromboembolic pulmonary hypertension after pulmonary embolism. N Engl J Med 2004; 350:2257–2264. 23. Trerotola SO, Lund GB, Scheel PJ Jr, Savader SJ, Venbrux AC, Osterman FA Jr. Thrombosed dialysis access grafts: percutaneous mechanical declotting without urokinase. Radiology 1994; 191:721–726. 24. Ribeiro A, Lindmarker P, Johnsson H, Juhlin-Dannfelt A, Jorfeldt L. Pulmonary embolism: a follow-up study of the relation between the degree of right ventricle overload and the extent of perfusion defects. J Intern Med 1999; 245:601–610. 25. Fisher MR, Forfia PR, Chamera E, et al. Accuracy of Doppler echocardiography in the hemodynamic assessment of pulmonary hypertension. Am J Respir Crit Care Med 2009; 179:615–621. 26. Litwin SE. Noninvasive assessment of pulmonary artery pressures: moving beyond tricuspid regurgitation velocities. Circ Cardiovasc Imaging 2010; 3:132–133.
