Glomerular filtration rate after left renal vein division and reconstruction during infrarenal aortic aneurysm repair

Methods 

A retrospective review was conducted on a prospectively compiled computerized database of all patients who underwent AAA repair at our center between January 2001 and March 2006. A total of 1189 patients underwent elective open repair of an infrarenal AAA. Patients with a medical history of previous renal parenchymal disease or surgery and those who had required proximal aortic clamping above the level of the renal arteries were excluded from the analysis to eliminate confounding factors.

Surgery was performed through a transperitoneal approach and intravenous heparin (70 IU/kg) was administered before aortic cross-clamping. Aortic reconstruction was performed with Dacron graft interposition, and proximal and distal anastomoses were made in an end-to-end fashion.

To gain proximal aortic control, LRV division was deemed necessary in 15 cases (1.3%); LRV was occluded in one patient. In the other 14 patients (LRV group), the LRV was reconstructed after the aortic anastomoses were completed and clamps were released. LRV reconstruction was done by direct reanastomosis in 10 cases and by interposition of an expanded polytetrafluoroethylene graft in four. Collateral branches were preserved during LRV mobilization and retraction. The control group was 56 patients without LRV division who where randomly identified from the database in a 4:1 ratio. Post hoc 1:1 case-matched analysis was also performed.

Patient groups were stratified by preoperative risk factors including diabetes, tobacco use, hypertension, and hyperlipidemia using a simplified grading system according to the Society for Vascular Surgery Suggested Reporting Standards,17 by perioperative cardiac risk, according to the Goldman revised cardiac risk index (RCRI),18 and by stages of chronic kidney disease, according to the National Kidney Foundation guidelines.14

GFR was estimated by using the Cockcroft-Gault equation14, 19: (140 −age) × weight/72 × serum creatinine (where age is in years, actual body weight is in kg, and serum creatinine is in mg/dL; for women, the equation is multiplied by 0.85). GFR values are expressed as mL/min/1.73 m2. Preoperative and postoperative GFR values and trend up to day 4 after surgery were compared between the two groups.

Morbidity and mortality were recorded. Renal dysfunction was defined as a rise in serum creatinine exceeding the baseline value by 30% and surpassing an absolute level of 2.0 mg/dL.20 The normal serum creatinine was defined as ≤1.5 mg/dL. Myocardial infarction was suggested by electrocardiographic changes and confirmed by elevation of cardiac enzymes, regardless of symptoms. Respiratory failure was defined as ventilator dependence of >72 hours, need for postoperative reintubation, clinical data or culture confirmation of pneumonia, or the need for tracheostomy.6 Ileus was defined as a delay in gut motility lasting for >72 hours after surgery. Operative times, blood loss, and hospital length of stay were also recorded and compared between groups. Follow-up data at 6 months were available in 12 patients who underwent LRV reconstruction.

Results were analyzed using the Fisher exact test or χ2 test for categoric data, unpaired t test, or Mann-Whitney test for continuous data, as appropriate. To investigate the effect of LRV division and reconstruction on GFR trend with time, two-way analysis of variance (ANOVA) for repeated measures was performed with one within-factor (time) and one between-factor (group). All analyses were run using SPSS/PC+ 15.0 statistical software (SPSS Inc, Chicago, Ill) for Windows (Microsoft, Redmond, Wash), except the ANOVA for repeated measures, which was calculated using SAS 8.02 software (SAS Institute Inc, Cary, NC).

Results 

Analysis of demographics and risk factors (Table II) showed that the two groups were overall comparable, except that AAAs were larger in the LRV group (71.4 ± 17.1 mm vs 56.0 ± 14.6 mm; P = .003), and preoperative GFR was significantly worse in the LRV group (65.3 ± 19.0 vs 82.8 ± 22.3; P = .009). No statistically significant difference was observed between the two groups for serum creatinine level (1.2 ± 0.4 vs 1.0 ± 0.3; P = .06).

Table II. Comparison of patient demographics and preoperative risk factors

	Characteristics	LRV n (%)	Controls n (%)	P⁎
	Overall	14	56
	Gender
	Male	13(92.8)	54(96.4)	.493
	Age (years)	72.2 ± 5.3	69.1 ± 7.5	.150
	AAA diameter (mm)	71.4 ± 17.1	56.0 ± 14.6	.003
	Pre-op SCr (mg/dL)	1.2 ± 0.4	1.0 ± 0.3	.056
	Pre-op GFR (mL/min/1.73 m2]	65.3 ± 19.0	82.8 ± 22.3	.009
	Stage†
	1 (GFR ≥90)	1(7.1)	21(37.5)	.003
	2 (GFR 60-89)	5(35.7)	26(46.4)
	3 (GFR 30-59)	8(57.2)	9(16.1)
	4 (GFR 15-29)	0(0.0)	0(0.0)
	5 (GFR <15 or dialysis)	0(0.0)	0(0.0)
	Body mass index
	Underweight (15-18.5)	0(0.0)	1(1.8)	.833
	Ideal (18.5-25)	3(21.4)	15(26.8)
	Overweight (25-30)	9(64.3)	29(51.8)
	Obese (>30)	2(14.3)	11(19.6)
	Revised cardiac risk index‡
	I (0.5%)	0(0.0)	0(0.0)	.867
	II (1.3%)	8(57.2)	35(62.5)
	III (3.6%)	5(35.7)	16(28.6)
	IV (9.1%)	1(7.1)	5(8.9)
	Hypertension
	0 (Diastolic <90 mm Hg)	3(21.5)	12(21.5)	.912
	1 (Easily controlled, single drug)	5(35.7)	18(32.1)
	2 (Requires 2 drugs)	5(35.7)	18(32.1)
	3 (>2 drugs or uncontrolled)	1(7.1)	8(14.3)
	Diabetes
	0 (None)	11(78.6)	49(87.5)	.669
	1 (Adult onset, no insulin)	3(21.4)	7(12.5)
	2 (Adult onset, insulin controlled)	0(0.0)	0(0.0)
	3 (Juvenile onset)	0(0.0)	0(0.0)
	Smoking
	0 (None; abstinence >10 yr)	4(28.6)	16(28.6)	.950
	1 (None; abstinence 1-10 yr)	6(42.8)	20(35.7)
	2 (<1 pack/day or abstinence >1yr)	2(14.3)	9(16.1)
	3 (Current 1 ≥ pack/day)	2(14.3)	11(19.6)
	Hyperlipidemia§
	0	9(64.3)	39(69.6)	.252
	1	3(21.4)	4(7.2)
	2	0(0.0)	0(0.0)
	3	2(14.3)	13(23.2)
	Pulmonary status∥
	0	7(50.0)	35(62.5)	.133
	1	4(28.6)	19(33.9)
	2	2(14.3)	1(1.8)
	3	1(7.1)	1(1.8)
	ASA score
	1	0(0.0)	0(0.0)	.692
	2	5(35.7)	26(46.4)
	3	8(57.2)	28(50.0)
	4	1(7.1)	2(3.6)
	5	0(0.0)	0(0.0)

AAA, Abdominal aortic aneurysm; SCr, serum creatinine; ASA, American Society of Anesthesiologists; GFR, = estimated glomerular filtration rate. Data are presented as n (%) for categoric variables, mean ± standard deviation for continuous variables, where appropriate.

⁎ Fisher exact test or χ2 test for categoric data, unpaired t test or Mann-Whitney test for continuous data. † 1, kidney damage, normal or ↑GFR; 2, kidney damage, mild ↓GFR; 3, moderate ↓GFR; 4, severe ↓GFR; 5, kidney failure. ‡ Major cardiac complication rates. § 0, cholesterol and triglycerides within normal limits for age; 1, mild elevation, diet controlled; 2, types II, III, or IV, strict diet control; 3, requires drug control. ∥ Pulmonary status: 0 = asymptomatic, normal chest x-ray film, pulmonary function tests (PFTs) 20%; 1, mild dyspnea on exertion or mild x-ray parenchymal changes, PFTs 65%-80%; 2 = between 1 and 3; 3, vital capacity <1.85 L, forced expiratory volume in 1 second <1.2 L or < 35%, Pco2 >45 mm Hg, pulmonary hypertension.

Results from multivariate analysis of variance (Fig 1) showed the main effect of group to be significant (F1,68 = 7.86; P = .007). The analysis did not reveal a significant effect for time (F4,65 = 0.79; P = .53) or for group × time interaction (F4,65 = 1.18; P = .33). The variation of GFR at day 4 after surgery compared with preoperative values was not statistically different (5.6 ± 12.6 vs 1.0 ± 15.5; P = .67). In addition, for each patient of the LRV group, a matched-control subject was selected among those of the control group by aneurysm size (SD of LRV group, ± 17 mm) and preoperative GFR (SD of LRV group, ±19). As expected, a further multivariate analysis of variance showed that the main effect of group was not significant (F1,26 = 0.24; P = .63), and consistent with the previous analysis, no significant effect for time (F4,23 = 0.14; P = .97) or group × time interaction (F4,23 = 1.88; P = .15) was found.

View Large Image Download to PowerPoint

Fig 1. Variations of glomerular filtration rate (GFR) values at different postoperative times. Within-factor effect (time): F4,65 = 0.79; P = .53; between-factor effect (group): F1,68 = 7.86; P = .007; group × time interaction: F4,65 = 1.18; P = .33.

Postoperative renal dysfunction was observed in two patients of LRV group and one of control group. The difference in operative time was not significant (148.4 ± 35.8 minutes vs 131.0 ± 40.3 minutes; P = .07). Two (14.3%) of 14 patients in the LRV group had an aortobiiliac reconstruction; in the control group, an aortobiiliac/bifemoral reconstruction was performed in nine (16.1%) of 56. No differences were noted for blood loss, perioperative complications, and hospital length of stay between groups (Table III). No deaths were recorded. A 6-month follow-up of renal function in 12 patients of LRV group showed no significant decrease in GFR compared with day 4 postoperative values (70.8 ± 24.8 vs 69.1 ± 23.5; P = .86). At duplex scan, the reconstructed LRV could be insonated in nine of these patients and was patent in all of them.

Table III. Comparison of perioperative variables

	Variable	Left renal vein n (%)	Controls n (%)	P⁎
	Total operative time (min)	148.4 ± 35.8	131.0 ± 40.3	.070
	Aortic cross-clamping time (min)	31.6 ± 11.3	32.4 ± 11.0	.849
	Blood loss (mL)	585.7 ± 264.2	567.7 ± 222.5	.877
	Hospital length of stay (days)	6.2 ± 1.8	5.5 ± 1.2	.103
	Major complications
	Total	5(35.7)	8(14.3)	.117
	Renal dysfunction	2(14.3)	1(1.8)
	Myocardial infarction	0(0.0)	0(0.0)
	Respiratory failure	3(21.4)	5(8.9)
	Ileus	0(0.0)	2(3.6)

Data are presented as n (%) for categoric variables, mean ± standard deviation for continuous variables, where appropriate.

⁎ Fisher exact test or χ2 test for categoric data, unpaired t test or Mann-Whitney test for continuous data.

Discussion 

The division of the LRV, initially reported as a surgical adjunct for retroperitoneal tumor resection21 and portorenal shunt,22 facilitates control of the pararenal aorta during AAA repair. From a technical standpoint, LRV collaterals, including inferior adrenal, phrenic, gonadal, and lumbar veins, need to be preserved. Caution should be exerted during retraction of the LRV to avoid injury to these branches. If the LRV is eventually divided after its collaterals are sacrificed, the risk of left kidney dysfunction is significantly increased.2

Acute complications of LRV ligation are rare and mainly reported in patients with a ruptured AAA. These include increased mortality, massive hemorrhage requiring nephrectomy, venous renal infarction, and loss of left renal function.10, 23, 24 During open surgical repair of pararenal aortic aneurysms, LRV ligation was also found to be an independent determinant of postoperative renal insufficiency and was also associated with cardiopulmonary complications and prolonged hospital stay.6

Among chronic sequelae, subclinical or clinical deterioration of renal function is the most common. In case of insufficient drainage through LRV tributaries, renal venous pressure significantly increases, resulting in a reduction of renal blood flow and GFR with activation of the renin-angiotensin-aldosterone system, which further decreases both renal perfusion and GFR.25 In addition, as described in the LRV entrapment or “nutcracker” syndrome,26, 27 which is also characterized by renal venous hypertension, proteinuria and hematuria may be present,4, 25 the latter as a consequence of the rupture of extensively developed perirenal and pararenal varicosities into the adjacent collecting system.28 Finally, an increase in LRV pressure may produce the pelvic congestion syndrome, which consists of dysmenorrhea, dyspareunia, dysuria, and vulvar and pelvic varices in women, and varicocele in men.27

Calligaro et al2 proposed intraoperative LRV stump pressure measurement after test clamping to predict the occurrence of venous hypertension. A stump pressure ≥50 cm H2O and a very distended LRV after clamping, regardless of the pressure, were defined as contraindications to LRV division.2 Absolute LRV pressure has been reported to be unreliable, however,28, 29 and a renal-caval pressure gradient of 1 to 3 mm Hg or greater was recommended instead to assess for LRV hypertension.28, 30 Also, Takebayashi et al28 demonstrated that a distended LRV does not necessary imply a significantly increased pressure gradient.

In our experience, LRV was divided during AAA repair in 1.3% of cases and its reconstruction routinely performed when the vein was patent. In contrast with previous reports,2, 3 this maneuver did not significantly lengthen operative time or was associated with increased complication rates.

Kidney function after LRV reanastomosis was assessed by GFR estimations using the Cockcroft-Gault formula, which represents one of the most accurate equations to estimate GFR.31 Our data showed that reconstitution of LRV anatomic continuity was associated with the maintenance of preoperative renal functional status. Transient postoperative GFR increase was observed in agreement with previous reports,9, 32 likely as a result of increased extracellular fluid volume.32

Of interest was that LRV division was associated with larger aneurysms and significantly lower preoperative GFR values. In these patients at higher risk for clinical or subclinical renal failure, no GFR reduction was observed after reanastomosis of the vein compared with patients in whom the LRV was left intact. Also, our findings are consistent with the relevance of AAA diameter as a marker of progressive cardiovascular disease33 and an independent predictor of cardiovascular mortality, excluding aneurysm-related death.33, 34 Finally, LRV reconstruction appears to be durable both in terms of vein patency and preservation of kidney function; however, our results at 6 months need to be confirmed by a long-term follow-up on a larger group of patients. Other limitations also include the lack of a control group of patients with LRV ligation and the use of GFR estimations related to the retrospective fashion of the study.

Conclusion 

Based on our experience, we can draw no thorough conclusions about the optimal management of the LRV after its division; that remains a matter of debate. If collateral circulation is adequate, LRV ligation may not adversely affect renal function in many cases; however, based on preoperative and intraoperative findings, the occurrence of renal venous hypertension seems to be unpredictable.2, 28, 29 In this respect, LRV reanastomosis and LRV transposition for the treatment of the nutcracker phenomenon are safe maneuvers that normalize venous renal circulation.35

Our study showed that GFR after LRV division and reconstruction during infrarenal AAA repair remains unchanged. Rather than being considered unnecessary or cumbersome in high-risk patients, restoration of anatomic continuity of LRV should be regarded as a feasible and safe procedure.

Author contributions