Left Renal Vein Division During Open Surgery of Abdominal Aortic Disease: A

Propensity Score-Matched CaseeControl Study

L. Wang, S.J. Xin *, Z. Song, J. Zhang
Division of Vascular and Thyroid Surgery, First Hospital of China Medical University, China

WHAT THIS PAPER ADDS

As management of the left renal vein in abdominal aortic surgery is still controversial, we conducted a pro-
pensity score-matched caseecontrol study to evaluate the effect of left renal vein division, and concluded that
left renal vein division did not increase the risks of early or late mortality and morbidity in patients who un-
derwent open surgery of the abdominal aorta.

Objective: To investigate the impact of left renal vein division (LRVD) on the postoperative renal function of
abdominal aortic aneurysm (AAA) or aortic occlusive disease (AOD).
Methods: Between January 2000 and January 2012, 238 patients, including 179 AAAs and 59 AODs underwent
open surgery in our institution (patients who required suprarenal aortic clamping were excluded). 49 patients (38
AAAs, 11 AODs) required LRVD during the operation. Patients in the LRVD and non-LRVD groups were matched
using propensity score method in a 1:2 ratio. Pre- and postoperative renal function, major complications, in-
hospital mortality and long-term renal function were compared. Late survival rate was estimated using the
KaplaneMeier method.
Results: Overall, the LRVD group had a significantly higher male/female ratio (39/10 vs.122/67, p ¼ .045), higher
ruptured AAA ratio (36.7% vs. 20.1%, p ¼ .014), higher American Society of Anesthesiologists (ASA) classification
3 (53.1% vs. 30.2%, p ¼ .003), higher co-morbidities of coronary artery disease (51.0% vs. 33.3%, p ¼ .022),
higher preoperative shock (22.4% vs. 8.5%, p ¼ .006) and longer operative time (164.2  43 vs. 150.1  41 min,
p ¼ .035). With propensity score matching (PSM), 48 patients in the LRVD group and 96 in the non-LRVD group
were enrolled in this study. The baseline characteristics were well balanced in the groups (p < .05) after PSM.
There were no statistically significant differences in preoperative glomerular filtration rate (GFR, expressed as
mL/min/1.73 m2) (62.0  13.1 vs. 62.9  12.9, p ¼ .695), and postoperative GFR on day 1 (60.3  13.7 vs.
61.3  13.1, p ¼ .671), day 3 (54.6  16.8 vs. 58.8  14.3, p ¼ .120), day 7 (62.1  16.8 vs. 63.7  13.4 p ¼ .537)
and in the long term (>12 months) (62.4  14.0 vs. 64.7  11.8 p ¼ .302). There were no statistically significant
differences in in-hospital mortality (6.3% vs. 9.2%, p ¼ .522) and late survival rate estimated by the Kaplane
Meier method (p ¼ .96).
Conclusion: LRVD may be a safe maneuver during abdominal aortic surgery as it did not increase the risks of early
or late mortality and morbidity.
Ó 2013 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved.
Article history: Received 7 December 2012, Accepted 21 April 2013, Available online 3 June 2013
Keywords: Aortic surgery, Left renal vein, Renal function

INTRODUCTION during a bleeding situation. On the other hand, LRVD may

To gain adequate exposure of the infrarenal aorta during also lead to venous hypertension and dysfunction of the left
surgery for abdominal aortic aneurysm (AAA) or aortic kidney. There are several published studies giving conflicting
occlusive disease (AOD), left renal vein division (LRVD) is results on the safety of LRVD.2e8
often required. According to the literature, LRVD rate More clinical data need to be accumulated, but it is
ranges from 1.3% to 18.8%.1,2 This adjunctive maneuver difficult to perform a large sample study in a single centre.
could facilitate the operation and sometimes save time And, because of the ethical problem, we cannot collect
evidence from a randomized clinical trial (RCT). In this study,
we made some amelioration in methodology. A propensity
* Corresponding author. S.J. Xin, Division of Vascular and Thyroid Surgery, score-matched (PSM) method, which could replicate some
Department of General Surgery, First Hospital of China Medical University, of the characteristics of a RCT,9 was used to control con-
155 Nanjing North Street, Shenyang 110001, China. founding factors and selective bias between LRVD patients
E-mail address: sjxin@mail.cmu.edu.cn (S.J. Xin).
1078-5884/$ e see front matter Ó 2013 European Society for Vascular
and non-LRVD patients who underwent abdominal aortic
Surgery. Published by Elsevier Ltd. All rights reserved. surgery. Both the early and long-term renal function were
http://dx.doi.org/10.1016/j.ejvs.2013.04.028 compared.
228 European Journal of Vascular and Endovascular Surgery Volume 46 Issue 2 August/2013

PATIENTS AND METHODS Pre- and postoperative (day 1, 3 and 7) renal function,
A retrospective review was carried out of all patients major complications, in-hospital mortality, late survival rate
diagnosed as AAA or AOD treated with open surgery at First and long-term renal function were compared. For renal
Hospital of China Medical University between January 2000 function evaluation, both serum creatinine (sCr) and
and January 2012. To avoid the confounding factors of renal glomerular filtration rate (GFR) were used. GFR was esti-
ischemia impact on the postoperative renal function, pa- mated by the Cockcroft-Gault equation:10,11 (140 e
tients requiring suprarenal aortic clamping were excluded age)  weight/72  sCr (age is in years, body weight is in kg
from this study. Two-hundred and thirty-eight (238) pa- and sCr is in mg/dL; the equation is multiplied by 0.85 in
tients, including 123 stable AAAs (sAAA) (51.7%), 56 women). The GFR values are expressed as mL/min/1.73 m2.
ruptured AAAs (rAAA) (23.5%) and 59 AODs (24.8%), were Major complications include myocardial infarction (MI),
eligible. Patients with CT and laparotomic findings of pulmonary embolism (PE), respiratory failure, renal
retroperitoneal hematoma or free peritoneal blood were dysfunction, major amputation, cerebral infarction and
diagnosed as rAAA. All rAAAs were treated urgently. ileus. MI was diagnosed by electrocardiographic change and
A transperitoneal midline incision was used in all cases. elevated cardiac enzymes. Respiratory failure was defined
Blood flow was reconstructed with a bifurcated PTFE graft as a patient who could not be independent from a venti-
in the form of end-to-end anastomy in proximal aorta and lator >72 hours, or required postoperative reintubation or
end-to-side anastomy in distal iliac arteries or femoral ar- tracheostomy. The diagnosis of PE or cerebral infarction was
teries. To improve the exposure of proximal neck, LRVD was made using radiographic evidence and consultation of a
performed in 49 patients (20.6%), including 20 sAAA, 18 senior specialist. Major amputation indicated below- or
rAAA and 11 AOD. All left renal veins were ligated and above-knee amputation for severe lower limb ischemia.
divided near the inferior vena cava to preserve collateral Renal dysfunction was defined as a rise in sCr >0.5 mg/dL
tributaries for drainage of the left kidney. There was no left or an absolute sCr level >2.0 mg/dL. Ileus was considered if
renal vein reconstruction in this group. the gut motility did not recover till 72 hours after surgery.1
Theoretically, patients who needed LRVD may have had We used the propensity score-matched method to bal-
worse baseline characteristics and surgical complexity ance the confounding factors between the LRVD and non-
compared with non-LRVD patients. In order to attenuate LRVD groups. For each patient, propensity score was
compounding variables, we conducted a propensity score- calculated using a multivariable logistic regression model
matched caseecontrol study. with the end point of LRVD (non-LRVD coded as 0, LRVD

Table 1. Patient demographics prior to and after propensity score-matching.

All patients Matched patients
LRVD Non-LRVD p Value LRVD Non-LRVD p Value
n ¼ 49 n ¼ 189 n ¼ 48 n ¼ 96
Mean age (years) 67 (51e84) 65 (49e82) .069 67 (51e84) 67 (54e82) .722
Male/female 39/10 122/67 .045 38/10 72/24 .579
Diagnosis n (%)
sAAA 20 (40.8%) 103 (54.5%) .087 20 (41.7%) 38 (39.6%) .810
rAAA 18 (36.7%) 38 (2.1%) .014 17 (35.4%) 33 (34.4%) .901
AOD 11 (22.4%) 48 (25.4%) .670 11 (22.9%) 25 (26.0%) .683
ASA classification n (%)
1 0 (.0%) 0 (.0%) 1.000 0 (.0%) 0 (.0%) 1.000
2 12 (24.5%) 109 (57.7%) <.001 12 (25.0%) 28 (29.2%) .599
3 26 (53.1%) 57 (30.2%) .003 26 (54.2%) 50 (52.1%) .813
4 10 (20.4%) 21 (11.1%) .085 10 (20.8%) 18 (18.8%) .766
5 1 (2.0%) 2 (2.0%) .583 0 (.0%) 0 (.0%) 1.000
Co-morbidities n (%)
Hypertension 32 (65.3%) 114 (60.3%) .523 32 (66.7%) 65 (66.3%) .900
CAD 25 (51.0%) 63 (33.3%) .022 24 (50%) 44 (45.8%) .637
DM 9 (18.4%) 34 (18.0%) .951 9 (18.8%) 22 (21.9%) .566
CVA 8 (16.3%) 21 (11.1%) .320 8 (16.7%) 13 (13.6%) .616
Renal insufficiency (Cr  1.5 mg/dL) 3 (6.1%) 5 (2.6%) .229 3 (6.3%) 4 (4.2%) .584
COPD 2 (4.1%) 3 (1.6%) .278 2 (4.2%) 2 (2.1%) .473
Shock n (%) 11 (22.4%) 16 (8.5%) .006 11 (22.4%) 14 (14.6%) .213
Operative time (min) 164.2  43 150.1  41 .035 164.8  43 159.1  40 .433
Aortic clamping time (min) 37.5  9 37.9  11 .815 37.6  9 38.7  10 .521
Intraoperative blood loss (mL) 577  220 526  248 .191 569  224 572  238 .942
AAA ¼ abdominal aortic aneurysm; AOD ¼ aortic occlusive disease; ASA ¼ American Society of Anesthesiologists; CAD ¼ coronary artery
disease; DM ¼ diabetes mellitus; COPD ¼ chronic obstructive pulmonary disease; CVA ¼ cerebrovascular accident; LRVD ¼ left renal vein
division.
L. Wang et al. 229

coded as 1). The covariates included age, sex, diagnosis, Table 3. Pre- and postoperative GFR (mean  SD), expressed as
preoperative shock (systolic blood pressure < 90 mmHg), mL/min/1.73 m2.
co-morbidities, American Society of Anesthesiologists (ASA) LRVD group Non-LRVD group p Value
classification, intraoperative blood lost, aortic clamping Preoperative 62.0  13.1 62.9  12.9 .695
time and operating time. Then the nearest neighbor GFR
matching was made in a 1:2 ratio. The matching caliper Postoperative 60.3  13.7 61.3  13.1 .671
width was 0.1. Qualitative variables were compared using a day 1GFR
Postoperative 54.6  16.8 58.8  14.3 .120
chi-square test and quantitative variables with an inde-
day 3 GFR
pendent sample t-test. Values of p <.05 were considered
Postoperative 62.1  16.8 63.7  13.4 .537
statistically significant. Late survival rate was established day 7 GFR
using KaplaneMeier analysis. Statistical analysis was per- Long-term 62.4  14.0 64.7  11.8 .302
formed by SPSS (version 12.0) for Windows (SPSS Inc, Chi- GFR
cago, IL, USA). GFR ¼ glomerular filtration rate; LRVD ¼ left renal vein division.
The study was approved by the ethics committee of
China Medical University. No patients required dialysis, or died directly related to
renal failure during follow-up (5e125 months, average 41
months).
RESULTS
A total of 144 patients were included in our study (48 in the DISCUSSION
LRVD group and 96 in the non-LRVD group). One rAAA Although the incidence of abdominal aortic disease is
patient in the LRVD group was not matched. Prior to increasing with the aging of Chinese population, the
matching, LRVD had significantly higher male/female ratio, amount of open surgery is decreasing. This is because of a
ruptured AAA ratio, ASA classification, co-morbidities of widely accepted endovascular therapy for AAA or AOD.
coronary artery disease, preoperative shock rate and longer However, increased complexity of open aortic surgery had
operative time. After matching, these variations were well been noticed.12 Cases left to open surgery were often in
balanced between groups. Detailed patient demographics emergency or with unfavorable anatomy, especially the
are shown in Table 1. short proximal neck. Thus, more LRVD was needed to
Both preoperative sCr and GFR showed no significant improve the exposure of the aorta. In our series, the ratio of
difference. We found a compromised renal function on LRVD was 20.6% in all and 21.2% in AAA patients, which
postoperative day 1 and 3, which recovered to the baseline was higher than previous reports.1,2
level on day 7. However, there was still no significant dif- Renal hypertension caused by LRVD resulted in a reduc-
ference on postoperative renal function. In 115 patients (37 tion of renal blood flow and GFR. The activation of the
in the LRVD group, 78 in the non-LRVD group) survived renineangiotensinealdosterone system may further reduce
more than 12 months, long-term renal function was ob- GFR.13 These pathophysiological changes were similar to
tained in an average of 31 months after surgery (13e78 the nutcracker syndrome, which was characterized by
months). Renal functions of six patients (four in the LRVD impeded outflow from the LRV into the inferior vena cava
group, two in the non-LRVD group) were failed to be as a result of extrinsic compression.14 To preserve venous
evaluated. Data available also showed no significant dif- collaterals and reduce the risk of stump thrombosis, all left
ference. Details of sCr and GFR are shown in Tables 2 and 3. renal veins were divided close to the inferior vena cava.
No significant difference was observed between groups Some early reports described complications of bleeding,
in major complications (Table 4). Total in-hospital mortality renal edema or renal rupture caused by division near the
was 8.3% (12 patients), all of these patients were rAAA left kidney.15e17 In our series, no patient experienced any
except for one sAAA who died from PE in the non-LRVD symptoms of nutcracker syndrome, such as left flank pain or
group. Late survival rate was similar between groups massive hematuria,14 both in-hospital and follow-up. There
(p ¼ .960). The KaplaneMeier curves are shown in Fig. 1.
Table 4. Comparison of major complications and in-hospital
Table 2. Pre- and postoperative sCr (mean  SD), expressed as survival rate.
mg/dL. Major complications LRVD group Non-LRVD group p Value
LRVD group Non-LRVD group p Value MI 2 (4.2%) 1 (1.0%) .216
Preoperative sCr .92  .27 .94  .25 .660 PE 0 (0%) 1 (1.0%) .478
Postoperative 1.06  .25 1.05  .26 .826 Respiratory failure 11 (22.9%) 17 (17.7%) .467
day 1 sCr Renal dysfunction 6 (12.5%) 11 (11.5%) .855
Postoperative 1.14  .34 1.07  .32 .228 Major amputation 2 (4.2%) 3 (3.1%) .748
day 3 sCr Cerebral infarction 1 (2.1%) 2 (2.1%) 1.000
Postoperative .97  .37 .96  .28 .857 Ileus 5 (10.4%) 12 (12.5%) .715
day 7 sCr In-hospital 3 (6.3%) 9 (9.2%) .522
Long-term sCr .95  .29 .93  .23 .653 mortality (%)
LRVD ¼ left renal LRVD ¼ left renal vein division; MI ¼ myocardial infarction;
vein division. PE ¼ pulmonary embolism.
230 European Journal of Vascular and Endovascular Surgery Volume 46 Issue 2 August/2013

Figure 1. The KaplaneMeier curves of surviving LRVD and non-LRVD patients (Cum Survival: %; Time: month). There was no significant
difference between groups (p ¼ .960 by log-rank test).

was also no symptomatic PE in the LRVD group. Re- requiring suprarenal aortic clamping were excluded, and the
anastomosis of the left renal vein might be necessary LRVD group was matched using the propensity score
when the collaterals, especially inferior adrenal and gonadal method to make the two groups more comparable.
veins, are in absence. There are still some limitations of this study. Post-
So far, the safety of LRVD without re-anastomosis is still operative imaging of computed tomography angiography or
controversial. Some studies showed that the LRVD did not ultrasonography may help us better understand the physi-
influence the in-hospital renal function or associated com- ological and pathological changes in the left kidney after
plications of aortic surgery.2e4 Samson and colleagues5 LRVD.18 Further determination of split renal function is
investigated the long-term renal function of 36 LRVD pa- recommended to reveal the direct impact of LRVD on the
tients, and found that all patients had a stable sCr and GFR left kidney.
level except for two, who had insufficient preoperative In conclusion, LRVD does not adversely affect the post-
renal function and experienced deterioration more than a operative renal function, complications and early or late
year after surgery. These findings demonstrated that LRVD survival rate of patients undergoing abdominal aortic sur-
was a safe adjunct for aortic surgery. gery. It is a safe procedure for complex or emergency aortic
Some other reports gave a conflicting conclusion, in that surgery. Reconstruction of the left renal vein may be un-
increased in-hospital sCr and decreased GFR after LRVD necessary when the collaterals are preserved.
were observed.6,7 AbuRahma6 reported that two of 13
LRVD patients had no left renal function six months after
CONFLICT OF INTEREST
surgery. In a multivariate analysis for open surgery of par-
arenal AAA, West et al.8 demonstrated that LRVD was a None.
significant predictor of pulmonary complications, post-
operative renal insufficiency and prolonged hospital stay.
FUNDING
Marrocco-Trischitta et al.,1 reporting on a group of AAAs
who underwent LRVD and reconstruction during open None.
repair, concluded that left renal vein reconstruction main-
tained renal function and was not associated with length-
REFERENCES
ening operative time and increasing complications.
However, re-anastomosis of the vein can only be completed 1 Marrocco-Trischitta MM, Melissano G, Kahlberg A, Setacci F,
safely and quickly by a well-trained surgeon. It was difficult Segreti S, Spelta S, et al. Glomerular filtration rate after left
renal vein division and reconstruction during infrarenal aortic
to promote the reconstruction in all hospitals.
aneurysm repair. J Vasc Surg 2007;45(3):481e6.
Available reports could not fully answer whether LRVD 2 Mehta T, Wade RG, Clarke JM. Is it safe to ligate the left renal
was a marker of complexity of surgical procedure or vein during open abdominal aortic aneurysm repair? Ann Vasc
whether LRVD itself led to those complications. Obviously, a Surg 2010;24(6):758e61.
RCT design would be unethical for this issue. Suprarenal 3 Komori K, Furuyama T, Maehara Y. Renal artery clamping and
aortic clamping might be another impact factor on renal left renal vein division during abdominal aortic aneurysm
function. Hence, in our retrospective study, patients repair. Eur J Vasc Endovasc Surg 2004;27:80e3.
L. Wang et al. 231

4 Elsharawy MA, Cheatle TR, Clarke JM, Colin JF. Effect of left 11 Cockcroft DW, Gault MH. Prediction of creatinine clearance
renal vein division during aortic surgery on renal function. Ann from serum creatinine. Nephron 1976;16:31e41.
R Coll Surg Engl 2000;82:417e20. 12 Costin JA, Watson DR, Duff SB, Edmonson-Holt A, Shaffer L,
5 Samson RH, Lepore Jr MR, Showalter DP, Nair DG, Lanoue JB. Blossom GB. Evaluation of the complexity of
Long-term safety of left renal vein division and ligation to open abdominal aneurysm repair in the era of endovascular
expedite complex abdominal aortic surgery. J Vasc Surg stent grafting. J Vasc Surg 2006;43:915e20 [discussion
2009;50(3):500e4 [discussion 504]. 920].
6 AbuRahma AF, Robinson PA, Boland JP, Lucente FC. The risk of 13 Doty JM, Saggi BH, Sugerman HJ, Blocher CR, Pin R, Fakhry I,
ligation of the left renal vein in resection of the abdominal et al. Effect of increased renal venous pressure on renal
aortic aneurysm. Surg Gynecol Obstet 1991;173:33e6. function. J Trauma 1999;47:1000e3.
7 Huber D, Harris JP, Walker PJ, May J, Tyrer P. Does division of 14 Kurklinsky AK, Rooke TW. Nutcracker phenomenon
the left renal vein during aortic surgery adversely affect renal and nutcracker syndrome. Mayo Clin Proc 2010 Jun;85(6):
function? Ann Vasc Surg 1991;5:74e9. 552e9.
8 West CA, Noel AA, Bower TC, Cherry Jr KJ, Gloviczki P, 15 Devine TJ, Scott DF, Mayers KA, King RB. Massive haemorrhage
Sullivan TM, et al. Factors affecting outcomes of open surgical caused by left renal vein ligation. Br J Surg 1980;67:594e5.
repair of pararenal aortic aneurysms: a 10-year experience. 16 Scher KS, Coil JA, Dawson DL, Wroczynski AF. The fate of
J Vasc Surg 2006;43:921e7 [discussion 927e8]. the left kidney after division of its vein. Am Surg 1984;50:
9 D’Agostino Jr RB. Propensity score methods for bias reduction 613e6.
in the comparison of a treatment to a non-randomized control 17 Swanson RJ, Carlson RE, Olcott C, Stoney RJ. Rupture of the left
group. Stat Med 1998 Oct 15;17(19):2265e81. kidney following renosplenic shunt. Surgery 1969;65:32e40.
10 Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney 18 Elsharawy MA, Moghazy KM. Fate of left kidney after left renal
function e measured and estimated glomerular filtration rate. vein division during management of aortic occlusive disease.
N Engl J Med 2006;354(23):2473e83. Int Angiol 2009;28(1):56e61.