Risk of Acute Kidney Injury after Percutaneous Pharmacomechanical Thrombectomy Using AngioJet in Venous and Arterial Thrombosis

Accepted Manuscript Risk of acute kidney injury after percutaneous pharmacomechanical thrombectomy using Angiojet in venous and arterial thrombosis Guillermo A. Escobar, MD, Dillon Burks, BS, Matthew R. Abate, MD, Mohammed F. Faramawi, MD, Ahsan T. Ali, MD, Lewis C. Lyons, MD, Mohammed M. Moursi, MD, Matthew R. Smeds, MD PII:

S0890-5096(17)30646-5

DOI:

10.1016/j.avsg.2016.12.018

Reference:

AVSG 3263

To appear in:

Annals of Vascular Surgery

Received Date: 15 April 2016 Revised Date:

8 July 2016

Accepted Date: 5 December 2016

Please cite this article as: Escobar GA, Burks D, Abate MR, Faramawi MF, Ali AT, Lyons LC, Moursi MM, Smeds MR, Risk of acute kidney injury after percutaneous pharmacomechanical thrombectomy using Angiojet in venous and arterial thrombosis, Annals of Vascular Surgery (2017), doi: 10.1016/ j.avsg.2016.12.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

1 2 3 Risk of acute kidney injury after percutaneous pharmacomechanical thrombectomy using

5

Angiojet in venous and arterial thrombosis

RI PT

4

6

SC

7 8

Guillermo A. Escobar, MD, Dillon Burks, BS, Matthew R. Abate, MD, Mohammed F.

10

Faramawi, MD, Ahsan T. Ali, MD, Lewis C. Lyons MD, Mohammed M. Moursi MD, and

11

Matthew R. Smeds, MD

M AN U

9

12

Corresponding Author:

16 17 18 19 20 21

Assistant Professor of Surgery, Vascular Surgery Division

EP

15

Guillermo A Escobar, MD

University of Arkansas for Medical Sciences

AC C

14

TE D

13

4301 W Markham St. #520-2 Little Rock, AR 72205-7199

Phone (501) 686-6176, Fax (501) 686-5328 [email protected]

22 23

1

ACCEPTED MANUSCRIPT

24 Disclosures

26

No author has a conflict of interest related to this research, and no external funding was obtained

27

to obtain or analyze this data.

RI PT

25

AC C

EP

TE D

M AN U

SC

28

2

ACCEPTED MANUSCRIPT

Abstract

30

Background: Percutaneous mechanical thrombectomy (PMT) is commonly used to treat acute

31

thrombotic syndromes. Angiojet (AJ) forcibly sprays fibrinolytics to fragment and aspirate

32

thrombus. It’s known to cause hemolysis and gross hematuria, yet potential consequences to

33

renal function after AJ remain unstudied. We sought to determine the risk of acute kidney injury

34

(AKI) after AJ when compared to other lysis techniques.

35

Methods and Results: We retrospectively reviewed patients treated with thrombolysis over 5

36

years. We identified those treated with Angiojet or catheter-directed thrombolysis (CDT).

37

Demographics, indications, procedures and laboratory values within 3 days were recorded. AKI

38

was defined as an increase >25% above the baseline creatinine within 72hrs of the procedure.

39

102 patients (52 AJ, 50 CDT) had no statistical difference in mean age (50 and 51), indication

40

(arterial thrombosis 65% and 88%), or baseline creatinine (0.9 mg/dL and 1.0 mg/dL),

41

respectively. AKI occurred in 15(29%) treated with AJ, versus 4(8%) of CDT(P=0.007).

TE D

M AN U

SC

RI PT

29

Similar numbers of AJ and CDT patients underwent additional open surgical procedures

42

(21% and 30%, respectively P=NS). Multivariable analysis demonstrated that the odds of AKI

44

were only increased by AJ (OR 8.2 P=0.004, 95% CI 1.98-34.17), open surgery (OR 5.4

45

P=0.013, 95% CI 1.43-20.17) or a >10% drop in hematocrit (OR 4.0 P=0.03, 95% CI 1.15-

46

14.25).

47

Conclusions: In our observational study, AJ is an independent risk factor for AKI. Concomitant

48

open surgery and drop in hematocrit also raise the odds of AKI. Renal injury after AJ is under-

49

reported in the literature, and may be related to hemolysis from the device.

AC C

EP

43

3

ACCEPTED MANUSCRIPT

50

Introduction Percutaneous pharmacomechanical thrombectomy (PMT) is a popular and useful tool for

52

rapid thrombus removal in acute thrombotic syndromes. It incorporates physical manipulation of

53

the thrombus with intravascular delivery of thrombolytic solutions to increase the speed and

54

efficiency of lysis. PMT using a variety of devices has been embraced for the treatment of both

55

arterial 1, 2 and venous 3-5 acute thrombotic syndromes. The evidence-guided role for PTM over

56

other techniques remains unclear, while it is recommended over simple, catheter-directed

57

thrombolysis (CDT) in venous thrombolysis in the guidelines published by the American Venous

58

Forum 6, while the American College of Chest physicians recommend using open surgical

59

thrombectomy over any type of thrombolysis for arterial occlusions7.

M AN U

SC

RI PT

51

The AngioJet Rheolytic Thrombectomy System (Possis Medical, Minneapolis, MN) is a

61

PMT device that uses a high-velocity spray designed to fragment clots, and is usually combined

62

with solutions containing plasminogen activators (lytics) as the spray. In addition, retrograde-

63

directed jets also create a negative pressure Bernoulli effect, which aspirates some of the

64

fragmented blood and clot 8. When compared to standard catheter-directed thrombolysis or high-

65

dose systemic infusions of fibrinolytics, Angiojet can potentially reduce the treatment time, total

66

volume of fibrinolytic, ICU and hospital lengths of stay, hospital costs and even lower bleeding

67

complications 9, 10. However, there has not been a randomized trial confirming these findings,

68

and in some series there may be no difference in treatment times.

EP

AC C

69

TE D

60

However, a consequence of intravascular high-pressure jet spray is not only clot

70

dissipation, but also significant destruction of red blood cells leading to hemolysis. All patients

71

treated by Angiojet manifest some degree of hematuria11, and gross hematuria is typically seen

72

(Figure 1). Practitioners generally manage this as a benign side-effect, and “hematuria” is not

4

ACCEPTED MANUSCRIPT

73

mentioned (or considered a complication) in clinical manuscripts using Angiojet. Despite this,

74

acute hemolysis is a well-established cause of renal failure in other scenarios, such as

75

paroxysmal nocturnal hemoglobinuria and failing mechanical heart valves 12, 13. Dukkipati et al. reported a case of acute kidney injury (AKI) after the use of Angiojet in a

RI PT

76

young patient with pulmonary embolism 13. Despite this report, we are not aware of any

78

subsequent publication reporting, or evaluating adverse effects after using this device on renal

79

function. Recently, the PEARL Angiojet registry mentioned the need for dialysis attributable to

80

Angiojet in 5% of the patients enrolled, but did not report the incidence of acute renal injury,

81

worsening renal function 14, nor offer a proposed cause for needing dialysis in these patients. We

82

sought to determine if patients undergoing Angiojet thrombolysis were at increased risk of acute

83

renal injury, when compared to non-Angiojet catheter-directed thrombolysis and identify other

84

contributory variables.

85

Methods:

M AN U

TE D

86

SC

77

With institutional review board approval, we retrospectively reviewed a prospectivelymaintained database of patients at our institution from 2007-2013 with procedural codes related

88

to thrombolysis (CPT 37201, 37187, 37209 and 75898), and/or Percutaneous Mechanical

89

Thrombectomy (CPT 37187). Consent was not obtained as there was less than minimal risk to

90

the patient, and no direct benefit to those enrolled was anticipated based on our findings. The

91

approach of treatment was up to the practitioner at the time of the intervention, and not

92

protocoled. We identified which patients were treated with Angiojet and collected demographics,

93

indications, laboratory values before and after the procedure (up to 3 days) and determined the

94

incidence of acute kidney injury (AKI). Chronic kidney disease was diagnosed as any patient

95

with that diagnosis documented by a nephrologist prior to our procedure, or anyone with a

AC C

EP

87

5

ACCEPTED MANUSCRIPT

96 97

baseline creatinine of 1.5mg/dL or greater. There are no published criteria for detecting acute kidney injury from hemolysis, therefore we used a clinical definition commonly attributed for contrast-induced nephropathy.

99

AKI was defined as an increase in creatinine (Cr) >25% of baseline within 72hrs15, 16 or an

RI PT

98

absolute increase >0.5mg/dL. Patients on dialysis before Angiojet, duplicated codes, or those

101

without lab values obtained before and 24-72hrs after treatment were excluded for analysis. We

102

used the “worst” laboratory results within this time frame for our analysis (eg. the lowest

103

hematocrit and highest creatinine) We chose to report creatinine rise as the measure for acute

104

kidney injury, rather than a decrease in GFR because the only variable in the GFR calculation

105

that would change in 72hrs would be the creatinine (age and weight would be unchanged). Thus,

106

with a rise in creatinine, there would be a proportionate drop in GFR regardless of which is

107

analyzed. All patients were exposed to 270mgI/mL Iodixanol (Visipaque, GE Healthcare,

108

Princeton NJ, USA) which is a hyposmolar, non-ionic iodinated contrast agent.

111

M AN U

TE D

110

We also collected the same data in patients treated with catheter-directed thrombolysis (CDT) alone, and compared these outcomes to the Angiojet (AJ) group. Statistical analysis was performed using GraphPad Prism 6 software (GraphPad, La Jolla,

EP

109

SC

100

CA) and Stata 12.1 (StataCorp LP, College Station, Texas). Pearson’s exact test was used for

113

univariate analysis, two-tailed T-test analysis was used to compare two groups as paired or

114

unpaired variables. Logistic regression was used to assess the relationship between AJ use and

115

AKI. Confidence intervals were considered at 95% and statistical significance was accepted with

116

a P<0.05. A Bonferroni correction was calculated for each variable that was assessed.

117 118

AC C

112

Biological systems have multiple risk factors that may act in combination to increase the odds of acute renal injury, and not just the one we are studying (Angiojet). In order to better

6

ACCEPTED MANUSCRIPT

evaluate the “direct” and “indirect” effects of our variables on developing acute kidney injury,

120

we wanted to evaluate them using another statistical approach to help “confirm” our findings. To

121

do this, we created a “causal model” using our variables (like a flow diagram) leading to acute

122

kidney injury using DAGitty software. This may help us re-evaluate if interactions appear among

123

variables that may not be identified with standard univariable and logistic regression analysis, or

124

correct our results if their significance was overestimated by not considering that one variable

125

was unduly influenced by the presence of another. Causal model analysis offers a visual and

126

statistical way to determine the minimal “adjustment sets” (statistical models) needed to estimate

127

the “total effect” of Angiojet on causing AKI and estimate the “direct effect” of any one variable

128

on developing AKI 17, 18.

129

Results:

130

102 patients who were treated with endovascular techniques for thrombotic syndromes,

M AN U

SC

RI PT

119

and had complete data were included for analysis. Of these, 52 were treated with Angiojet and 50

132

with traditional CDT. Table 1 describes the pre-procedure (baseline) characteristics in each

133

group, none of which were significantly different. Creatinine kinase and myoglobin levels were

134

not collected before and after treatment in most of the patients studied and could not be included

135

for analysis. The average age was 50 (range 20-87, median 49), and the average baseline

136

creatinine was similar. The average hematocrit (HCT) before treatment was similar between both

137

groups. Arterial thrombolysis was the indication in 34 (65%) of the AJ group, and 44 (88%) in

138

the CDT group, and the rest were venous. Only 4 AJ patients (8%) and 8 CDT (16%) had a

139

baseline Cr >1.4mg/dL before the procedures.

140

AC C

EP

TE D

131

Higher incidence of AKI after AJ versus CDI

7

ACCEPTED MANUSCRIPT

141

Acute kidney injury occurred in 15/52 (29%) patients treated with Angiojet (AKI group) which accounted for 79% of all cases of AKI, while only 4/50 (8%) of the CDT had AKI

143

(P=0.007) (Figure 2). The mean absolute rise in creatinine in the AKI group was 0.5mg/dL,

144

which was on average 61% higher than their baseline, p=0.003. This was not explained by pre-

145

operative chronic renal failure as the baseline Cr before AJ in the patients who developed AKI

146

was similar to that in the AJ patients who did not get AKI (P=0.1). Treatment of arterial

147

thrombosis was the indication in 16 (84%) of all AKI cases. Two patients (both in AJ group)

148

required dialysis within 2 days of their procedure, and died during their hospitalization (the

149

causes of death were mesenteric ischemia and pulmonary embolus).

150

Open surgery increases odds of AKI, especially in AJ

151

Of 26 patients that had open surgery during the study period, a similar number of patients

152

were from the Angiojet and CDT groups, 11(42%) and 15(58%), respectively. The AJ group had

153

14 procedures, while 20 procedures were done in the CDT group (Table 2). Most of the

154

procedures in the CDT group were open thromboembolectomy (9), and five had fasciotomies,

155

four had a bypass or endarterectomy and only two had a major amputation. In the AJ group, five

156

patients each had fasciotomies and/or thromboembolectomy, while three had a bypass or

157

endarterectomy and there was only one major amputation. There was no statistically significant

158

difference in the incidence of these surgeries in each lysis group.

SC

M AN U

TE D

EP

AC C

159

RI PT

142

Nine (35%) of the surgery patients developed AKI, 8 of these were treated with Angiojet

160

and 1 from the CDT group. Of the AJ patients who developed AKI, their average age was 54

161

(range 36-74, median 46), 56% were male, 2 had diabetes, and three had a baseline Cr higher

162

than 1.4mg/dL. Of Angiojet patients that underwent surgery, 53% (8/15) had AKI, while only

163

9% (1/11) of CDT (non-AJ) that had surgery suffered AKI. This interaction was best

8

ACCEPTED MANUSCRIPT

164

demonstrated in figure 3 where those that underwent surgery and AJ had the largest rise in Cr

165

(P=0.001) compared to any other group. Drop in Hematocrit (HCT) >10% independently increases odds of AKI

167

Comparing the absolute post procedure hematocrit between the AJ and CDT groups

RI PT

166

demonstrated significantly lower levels in the AJ group (p=0.004), although the absolute

169

difference between both groups (4%) is small (mean HCT 32% vs 36%, Median 32% vs 32%,

170

respectively). However, when we evaluated blood loss as a function of the percent change from

171

baseline in the Angiojet vs. non-Angiojet controls, there was not only a statistically significant

172

difference, but also a more clinically relevant change. The CDT group decreased by only 5% of

173

their baseline as opposed to the Angiojet group, which decreased by 12% (P=0.009, 95% CI of

174

1.8 to 11.9), and the increase in odds was even more apparent when we looked at the change in

175

hematocrit in patients that did or did not have AKI (Figure 4). Based on this finding, we

176

analyzed the odds ratio of AKI after a drop in hematocrit of at least 10% in our univariable and

177

multivariable modeling.

TE D

M AN U

SC

168

Univariate analysis is presented in figure 5 and revealed that treatment with AJ, open

179

surgery or a drop in hematocrit of at least 10% significantly increased the odds ratio of AKI.

180

Multivariable logistic regression analysis shown in figure 6 confirmed these three variables as

181

significant risk factors that raise the odds of AKI (OR 8.22, P=0.004, 95% CI 1.98-34.17 for

182

Angiojet, OR 5.38 P=0.013 CI 1.43-20.17 for surgery and OR 4.04 P=0.03, 95% CI 1.17-14.25

183

for hematocrit drop of 10% or more). We additionally evaluated if the indication for the

184

procedure (arterial or venous thrombosis) influenced the odds of developing AKI, we did not

185

find any statistically significant risk (OR 0.4446706, P=0.303, 95% CI 0.095-2.08).

AC C

EP

178

9

ACCEPTED MANUSCRIPT

186

In order to strengthen our findings above of Angiojet on the development of AKI, we analyzed the odds ratio of Angiojet leading to AKI using a directed acyclic graph which

188

identified their potential interactions (Figure 7). DAGitty analysis of the diagram determined that

189

to estimate the total effect of AJ on AKI required statistical analysis of the following groups of

190

variables: age and arterial indication, age and chronic renal failure (CRF) and diabetes (DM),

191

arterial indication versus venous indication, CRF and DM and venous indication. To determine

192

the direct effect of AJ and the risk of developing AKI, statistical models should also include age,

193

indication for procedure, hematocrit drop and surgery; as each of these could independently

194

cause AKI. Individual statistical modeling for each of these groups of variables was not

195

statistically significant (data available upon request) and our findings were unchanged after DAG

196

analysis.

SC

M AN U

197

RI PT

187

Thus, AJ remained as a statistically significant risk factor for developing AKI, with an odds ratio which was greater than 8; and both surgery and a drop in hematrocrit >10% from

199

baseline increase the odds ratio for AKI in a statistically significant manner.

200 Discussion

EP

201

TE D

198

In our institution, we found that the use of Angiojet for thrombolysis significantly

203

increased the odds of developing AKI by a factor of eight. This large difference was independent

204

of other comorbidities that are traditionally associated with AKI. AKI occurred in AJ patients as

205

young as 36 years old, and only two of the AJ patients in the AKI group had diabetes.

206

AC C

202

While the final hematocrit was similar in both groups, we were surprised to find that the

207

change in hematocrit after the procedure independently increased the risk of AKI. It was also

208

remarkable that patients treated with Angiojet who developed AKI had the largest drop in

10

ACCEPTED MANUSCRIPT

hematocrit compared to all other groups of patients, even when accounting for surgical

210

procedures (figure 4). Of note, the hematocrit in this group averaged 29% (hemoglobin of

211

~9g/dL) so these patients were certainly not profoundly anemic during their treatment period,

212

and this anemia would not pass the typical threshold for transfusion (and was similar in both

213

groups). This implies that anemia alone is unlikely the cause for renal injury. If the drop in

214

hemoglobin was only from iatrogenic dilution, then the creatinine would have also fallen in

215

tandem and thus lowered the likelihood of diagnosing AKI (in which case we would be

216

underestimating the incidence of AKI by having lower creatinine values). Thus, we speculate

217

that the larger hematocrit drop in the AJ patients may be a marker of hemolysis from the device,

218

which in turn may be contributing to the higher incidence of renal injury in those treated with AJ.

SC

M AN U

219

RI PT

209

Renal failure from acute hemolysis seems to occur after one of two different scenarios. Free serum hemoglobin normally binds to plasma haptoglobin, but when this becomes saturated,

221

the remaining free hemoglobin forms a dimer which is filtered into the renal tubules.

222

Hemoglobin is reabsorbed and dissociates intracellularly to heme and globin, which is cytotoxic.

223

The other renally-caustic mechanism is that heme can bind to Tamm-Horsefall proteins in the

224

renal tubules which forms intratubular casts that physically occlude tubular flow, leading to

225

oliguria and azotemia 19, 20.

EP

To our knowledge, there have been no in-vivo studies in the effects of Angiojet on renal

AC C

226

TE D

220

227

function. Lin et al. evaluated vessel injury in porcine models with Angiojet with or without

228

thrombolysis solutions, but they did not study the effects on renal function or effects of

229

hemolysis 21. The CaVenT registry reported the outcomes of the use of thrombolysis (including

230

Angiojet) for venous occlusions, but did not include the results of renal function, even in its

231

long-term outcomes publication 5, 22. The large, multicenter, randomized ATTRACT trial also

11

ACCEPTED MANUSCRIPT

does not seem to be including renal function in its data collection 23. A report evaluating the

233

management of thromboembolic complications after angioplasty used Angiojet in 14/18 cases.

234

They stated that “No hemoglobinuria or renal dysfunction or failure was noted” 24, but it is not

235

clear in the manuscript if this was referring only in the three patients in the series that died, or

236

their study population in toto; especially given that most patients treated with Angiojet have

237

some degree of hemoglobinuria. The recently published PEARL registry assessing 283 patients

238

with acute limb ischemia treated with AJ found 5% developed renal failure requiring dialysis

239

directly attributable to AJ. Dialysis dependence was permanent in ¾ of these patients. We had 2

240

patients require dialysis after AJ (4%) which although was a similar incidence, a direct

241

correlation of AJ in our cases was not possible.

SC

M AN U

242

RI PT

232

Acute renal injury is important to measure and detect, even if no dialysis is needed acutely. Overwhelming intravascular hemolysis from various causes can result in acute tubular

244

necrosis and fulminant renal failure, but it has also been found that long-standing exposures to

245

lower degrees of hemolysis can lead to hemosiderin depositing in the kidneys, ultimately leading

246

to renal failure 25. The odds ratio of death or intracranial hemorrhage after thrombolysis is almost

247

2.5 if the patient develops renal failure 26. This implies that every renal injury counts, and even

248

when the patient’s creatinine does not seem to change initially, the damage to the kidney may be

249

cumulative 27. Thus, patients who require thrombolysis for restoration of arterial blood flow may

250

be at increased risk, because re-occlusion is common 2 leading to repeated therapy with Angiojet.

EP

AC C

251

TE D

243

We acknowledge several limitations to our study. We had limited numbers of patients

252

treated with Angiojet, although our series is comparable or larger than most published datasets

253

that evaluated the device’s efficacy for clot resolution in comparison to CDT. We could not

254

control for contrast-induced nephropathy (CIN) as a potential cause for AKI because the volume

12

ACCEPTED MANUSCRIPT

was not always included in the operative reports, and was not available to us in the data

256

extraction model we had access to. On the other hand, at our institution we only use 270mgI/mL

257

Iodixanol for intravascular procedures; and this iodinated contrast agent has the lowest reported

258

incidence of CIN given its hypo-osmolar, non-ionic properties. We are also encouraged that this

259

alone is unlikely to account for the large difference, as we studied the outcomes in two groups of

260

patients who were both treated with endovascular techniques that required exposure to iodinated

261

contrast agent. In addition, Angiojet should theoretically reduce contrast exposure by reducing

262

the treatment times and the number of contrast boluses needed for the repeated “lysis checks”

263

needed in traditional catheter-directed thrombolysis; thus we do not think that contrast exposure

264

alone could account for our large difference in odds ratio by multivariable and DAG statistical

265

analysis. Finally, rhabdomyolysis from ischemia/reperfusion in some of these cases could

266

certainly contribute to a risk of AKI, especially as nearly all of the patients with creatinine

267

elevation were treated for arterial thrombus. Unfortunately, only a few patients had myoglobin or

268

creatinine kinase levels drawn before and after the lysis procedure. However, rhabdomyolysis

269

alone is also unlikely to explain the odds ratio greater than 8 in the AJ group, given that almost

270

90% of the control (CDT group) was treated for arterial thrombosis (compared to only 65% of

271

AJ), and few patients required amputations or fasciotomies in either group. When we analyzed

272

the indication of the procedure (venous thrombolysis or arterial) as a risk factor, we did not find

273

any statistically significant difference in the odds of developing AKI, although this is not

274

definitive given the smaller numbers involved. If significant rhabdomyolysis alone were the

275

predominate cause of this difference, then we would have expected more patients would have

276

either had fasciotomies, or subsequent amputations in the AKI group, which was not found.

AC C

EP

TE D

M AN U

SC

RI PT

255

13

ACCEPTED MANUSCRIPT

277

Nevertheless, we certainly acknowledge that we cannot exclude this as a contributing factor in

278

this retrospective review.

279

In summary, AJ significantly increases the odds of renal injury when compared to CDT and is known to cause hemolysis. AJ has been directly linked to dialysis-dependent renal failure

281

in at least one large registry and in one published case report. We believe that hemolysis is not

282

benign to the kidney either acutely or chronically, and patients treated for recurrent thrombotic

283

syndromes may be at higher risk to both. Further study being done at our institution to better

284

identify other causal risk factors (Angiojet treatment times, time intervals between treatments,

285

duration of hematuria post-procedure, improved quantification of blood loss etc.) so as to better

286

define the ideal patients and techniques to optimally use AJ for acute intravascular thrombosis.

M AN U

SC

RI PT

280

287

AC C

EP

TE D

288

14

ACCEPTED MANUSCRIPT

References

290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333

1. Valji K, Roberts AC, Davis GB, Bookstein JJ. Pulsed-spray thrombolysis of arterial and bypass graft occlusions. AJR Am J Roentgenol. 1991;156(3):617-21. 2. Comerota AJ, Weaver FA, Hosking JD, Froehlich J, Folander H, Sussman B, et al. Results of a prospective, randomized trial of surgery versus thrombolysis for occluded lower extremity bypass grafts. Am J Surg. 1996;172(2):105-12. 3. Lin PH, Zhou W, Dardik A, Mussa F, Kougias P, Hedayati N, et al. Catheter-direct thrombolysis versus pharmacomechanical thrombectomy for treatment of symptomatic lower extremity deep venous thrombosis. Am J Surg. 2006;192(6):782-8. 4. Comerota AJ, Throm RC, Mathias SD, Haughton S, Mewissen M. Catheter-directed thrombolysis for iliofemoral deep venous thrombosis improves health-related quality of life. J Vasc Surg. 2000;32(1):130-7. 5. Enden T, Haig Y, Klow NE, Slagsvold CE, Sandvik L, Ghanima W, et al. Long-term outcome after additional catheter-directed thrombolysis versus standard treatment for acute iliofemoral deep vein thrombosis (the CaVenT study): a randomised controlled trial. Lancet. 2012;379(9810):31-8. 6. Meissner MH, Gloviczki P, Comerota AJ, Dalsing MC, Eklof BG, Gillespie DL, et al. Early thrombus removal strategies for acute deep venous thrombosis: clinical practice guidelines of the Society for Vascular Surgery and the American Venous Forum. J Vasc Surg. 2012;55(5):1449-62. 7. Alonso-Coello P, Bellmunt S, McGorrian C, Anand SS, Guzman R, Criqui MH, et al. Antithrombotic therapy in peripheral artery disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):e669S-90S. 8. Sharafuddin MJ, Hicks ME. Current status of percutaneous mechanical thrombectomy. Part I. General principles. J Vasc Interv Radiol. 1997;8(6):911-21. 9. Allie DE, Hebert CJ, Lirtzman MD, Wyatt CH, Keller VA, Khan MH, et al. Novel simultaneous combination chemical thrombolysis/rheolytic thrombectomy therapy for acute critical limb ischemia: the power-pulse spray technique. Catheter Cardiovasc Interv. 2004;63(4):512-22. 10. Bush RL, Lin PH, Bates JT, Mureebe L, Zhou W, Lumsden AB. Pharmacomechanical thrombectomy for treatment of symptomatic lower extremity deep venous thrombosis: safety and feasibility study. J Vasc Surg. 2004;40(5):965-70. 11. Ouriel K. Endovascular techniques in the treatment of acute limb ischemia: thrombolytic agents, trials, and percutaneous mechanical thrombectomy techniques. Semin Vasc Surg. 2003;16(4):270-9. 12. Concepcion B, Korbet SM, Schwartz MM. Intravascular hemolysis and acute renal failure after mitral and aortic valve repair. Am J Kidney Dis. 2008;52(5):1010-5. 13. Dukkipati R, Yang EH, Adler S, Vintch J. Acute kidney injury caused by intravascular hemolysis after mechanical thrombectomy. Nat Clin Pract Nephrol. 2009;5(2):112-6. 14. Leung DA, Blitz LR, Nelson T, Amin A, Soukas PA, Nanjundappa A, et al. Rheolytic Pharmacomechanical Thrombectomy for the Management of Acute Limb Ischemia: Results From the PEARL Registry. J Endovasc Ther. 2015;22(4):546-57. 15. Rezaei Y, Khademvatani K, Rahimi B, Khoshfetrat M, Arjmand N, SeyyedMohammadzad MH. Short-Term High-Dose Vitamin E to Prevent Contrast Medium-Induced

AC C

EP

TE D

M AN U

SC

RI PT

289

15

ACCEPTED MANUSCRIPT

371

EP

TE D

M AN U

SC

RI PT

Acute Kidney Injury in Patients With Chronic Kidney Disease Undergoing Elective Coronary Angiography: A Randomized Placebo-Controlled Trial. J Am Heart Assoc. 2016;5(3). 16. Centola M, Lucreziotti S, Salerno-Uriarte D, Sponzilli C, Ferrante G, Acquaviva R, et al. A comparison between two different definitions of contrast-induced acute kidney injury in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention. Int J Cardiol. 2016;210:4-9. 17. Textor J, Hardt J, Knuppel S. DAGitty: a graphical tool for analyzing causal diagrams. Epidemiology. 2011;22(5):745. 18. Textor J. Daggity Software. 2013 [cited 2015]; 2.1:[Available from: http://www.dagitty.net. 19. Qian Q, Nath KA, Wu Y, Daoud TM, Sethi S. Hemolysis and acute kidney failure. Am J Kidney Dis. 2010;56(4):780-4. 20. Nath KA, Grande JP, Croatt AJ, Likely S, Hebbel RP, Enright H. Intracellular targets in heme protein-induced renal injury. Kidney Int. 1998;53(1):100-11. 21. Lin PH, Mussa FF, Hedayati N, Naoum JJ, Zhou W, Yao Q, et al. Comparison of AngioJet rheolytic pharmacomechanical thrombectomy versus AngioJet rheolytic thrombectomy in a porcine peripheral arterial model. World J Surg. 2007;31(4):715-22. 22. Enden T, Sandvik L, Klow NE, Hafsahl G, Holme PA, Holmen LO, et al. Catheterdirected Venous Thrombolysis in acute iliofemoral vein thrombosis--the CaVenT study: rationale and design of a multicenter, randomized, controlled, clinical trial (NCT00251771). Am Heart J. 2007;154(5):808-14. 23. Comerota AJ. The ATTRACT trial: rationale for early intervention for iliofemoral DVT. Perspect Vasc Surg Endovasc Ther. 2009;21(4):221-4; quiz 4-5. 24. Spiliopoulos S, Katsanos K, Fragkos G, Karnabatidis D, Siablis D. Treatment of infrainguinal thromboembolic complications during peripheral endovascular procedures with AngioJet rheolytic thrombectomy, intraoperative thrombolysis, and selective stenting. J Vasc Surg. 2012;56(5):1308-16. 25. Ackermann D, Vogt B, Gugger M, Marti HP. Renal haemosiderosis: an unusual presentation of acute renal failure in a patient following heart valve prosthesis. Nephrol Dial Transplant. 2004;19(10):2682-3. 26. Bashir R, Zack CJ, Zhao H, Comerota AJ, Bove AA. Comparative outcomes of catheterdirected thrombolysis plus anticoagulation vs anticoagulation alone to treat lower-extremity proximal deep vein thrombosis. JAMA Intern Med. 2014;174(9):1494-501. 27. Nath KA, Croatt AJ, Haggard JJ, Grande JP. Renal response to repetitive exposure to heme proteins: chronic injury induced by an acute insult. Kidney Int. 2000;57(6):2423-33.

AC C

334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370

16

ACCEPTED MANUSCRIPT

372

Figure 1

373

A) Panel A (left) shows gross hematuria being collected in a urine collection tubing

374

B) Demonstrates the progression of hematuria from normal, clear urine prior to Angiojet (black arrow), followed by the progressively denser hematuria (white arrow).

RI PT

375 376

Figure 2

378

Raw incidence of acute kidney injury (AKI) defined as an increase greater than 25% of baseline.

379

(Angiojet = AJ, Catheter directed thrombolysis = CDT, Creatinine= Cr)

SC

377

M AN U

380 Figure 3

382

Percent change in the creatinine in treated patients when exposed to additional open surgery.

383

Dotted horizontal line indicates the threshold for definition of acute kidney injury (25% rise after

384

the procedure). (Angiojet = AJ, Catheter directed thrombolysis =CDT)

385 386 387

EP

TE D

381

Figure 4

389

Graph showing the effect of a drop in the hematocrit (as a percentage of baseline) after being

390

treated with Angiojet (AJ) or catheter directed thrombolysis (CDT), in those that did, or did not

391

have acute kidney injury (AKI).

392 393

AC C

388

Figure 5

17

ACCEPTED MANUSCRIPT

Univariable analysis and forest plot to demonstrate Odds Ratios (OR) and 95% confidence

395

intervals (CI) of variables expected to be related to development of AKI. Odds ratio axis is at a

396

Log 5 scale. (Angiojet = AJ, Major open surgical procedures = Maj Surg, Chronic renal failure =

397

CRF, patients with at least a 10% drop in their hematocrit = 10% Drop HCT)

RI PT

394

398 Figure 6

400

Multivariable logistic regression analysis and forest plot to demonstrate Odds Ratios (OR) and

401

95% confidence intervals (CI) of variables expected to be related to development of AKI. Odds

402

ratio axis is at a Log 5 scale. (Angiojet = AJ, Major open surgical procedures = Maj Surg,

403

Chronic renal failure = CRF, patients with at least a 10% drop in their hematocrit = 10% Drop

404

HCT)

M AN U

SC

399

405 Figure 7

407

Directed acyclic graphs (DAGs) using DAGitty software to determine the potential effects of

408

each variable. The graph was created by the authors using background knowledge of potential

409

interactions between the variables involved with lysis and developing AKI.

410

Lysis procedure is the primary exposure and acute kidney injury (AKI) is the outcome. Red

411

variables are potential confounders to the “total effect” of the lysis procedure on AKI, as they

412

potentially influence both the primary exposure and AKI. Blue variables are those that may bias

413

the “direct effect” of lysis procedure as they alone can cause AKI independently of the lysis

414

procedure. Green variables are determinants to the Lysis and do not directly affect AKI. (Arterial

415

thrombosis = Arterial_clot, Venous thrombosis = venous clot, Chronic renal failure = CRF,

416

Diabetes mellitus = DM, Drop in hematocrit = Drop_Hematocrit)

AC C

EP

TE D

406

18

ACCEPTED MANUSCRIPT

Table 1

418

Patient baseline demographics comorbidities and lab values for angiojet patients and catheter-

419

directed thrombolysis (CDT) patients. (Cr=Creatinine mg/dL, HCT= hematocrit, not significant

420

= NS).

421

Table 2

422

Adjunct endovascular and surgical procedures done to each group

SC

RI PT

417

AC C

EP

TE D

M AN U

423

19

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT