Lower Incidence of Postoperative Acute Kidney Injury in Robot-Assisted Partial Nephrectomy Than in Open Partial Nephrectomy: A Propensity Score-Matched Study

Abstract

Background:

Acute kidney injury (AKI) after partial nephrectomy is attributed to parenchymal reduction and ischemia, but the extent of its effect remains unclear. This study aimed to compare the incidence of postoperative AKI among surgical modalities, robot-assisted partial nephrectomy (RAPN), laparoscopic partial nephrectomy (LPN), and open partial nephrectomy (OPN), and to evaluate the validity of RAPN by comparing it with LPN and OPN in terms of postoperative AKI, perioperative complications, and long-term renal function.

Patients and Methods:

Patients who underwent RAPN, LPN, and OPN for renal tumors at our institutions between 2004 and 2018 were retrospectively analyzed. RAPN and LPN were performed under warm ischemia and OPN under cold ischemia. En bloc hilar clamping was employed for LPN and OPN and arterial clamping for RAPN. AKI was defined as % decrease in estimated glomerular filtration rate (eGFR) >25% from preoperative eGFR to postoperative nadir eGFR. Multivariate regression analysis was used to test associations of AKI with perioperative factors. Then, we compared the incidence of AKI with two propensity score-matched cohorts: RAPN vs OPN and RAPN vs LPN.

Results:

This study included 1762 cases (RAPN: 959, LPN: 215, and OPN: 588). After matching, 147 cases each from RAPN and LPN groups and 368 cases each from RAPN and OPN groups were selected. RAPN had shorter warm ischemia time than LPN, lower incidence of AKI, and lower % decrease in eGFR after 6 months. RAPN had a shorter ischemia time and a lower incidence of AKI than OPN, although the % decrease in eGFR after 6 months did not differ significantly.

Conclusions:

AKI incidence was lower in RAPN than in LPN or OPN, which may be due to the shorter ischemia time or clamping of only arteries in RAPN. Although long-term renal outcomes did not differ between RAPN and OPN, RAPN can help prevent AKI. This supports the validity of RAPN for patients with chronic kidney disease.

Introduction

Partial nephrectomy (PN) is the reference standard of care for small renal masses and is preferred over radical nephrectomy because of its equivalent oncological outcomes and superior renal functional preservation.^1
–3 With the prevalence of minimally invasive surgery, laparoscopic partial nephrectomy (LPN) became common. Furthermore, robot-assisted partial nephrectomy (RAPN) has been increasingly employed because it enables precise tumor resection and reconstruction due to the extended range of motion of the da Vinci forceps.⁴

Acute kidney injury (AKI) after PN was reportedly due to the preoperative renal parenchymal quality and intraoperative parenchymal volume reduction and ischemic insult. However, the extent of each effect remains unclear.^5,6 Although warm ischemia time during PN is recommended as <25–30 minutes, which is shorter than that under cold ischemia, the ischemic insult with limited warm ischemia time can recover and may not affect long-term renal function.^6

–13

Thus, we aimed to compare the incidence of AKI among surgical modalities, namely RAPN, LPN, and open partial nephrectomy (OPN), and its effect on long-term renal function to evaluate the validity of RAPN.

Patients and Methods

Patient population

Institutional review board approval was obtained for the retrospective analysis of patient data (5107). Overall, 2068 patients who underwent PN for renal tumors at two high-volume institutions between January 2004 and December 2018 were enrolled. Among the excluded patients, 6 were <20 years old, 14 underwent PN with other simultaneous surgical procedures, 8 underwent PN with autotransplantation, and 204 had missing data. To decrease bias from the technical skills of each surgeon, data of surgeons without experience in all three surgical modalities (N = 74) were excluded. Finally, 1762 patients (OPN: 588, LPN: 215, and RAPN: 959) were retrospectively analyzed. The clinical and pathological parameters of each patient were obtained retrospectively from electronic medical charts. AKI was defined as the decrease in estimated glomerular filtration rate (eGFR) by >25% from preoperative eGFR to postoperative nadir eGFR within 30 days, according to the risk, injury, failure, loss, and end-stage kidney disease (RIFLE) criteria.¹⁴ First, multivariate logistic regression analysis was performed to test the correlation between the incidence of postoperative AKI and preoperative variables. Second, we assigned two matched cohorts (OPN vs RAPN and LPN vs RAPN) in which baseline patient and tumor characteristics were adjusted; then, we compared the renal functional outcomes and perioperative complication rate. Preoperative and postoperative global renal functions were assessed using eGFR before surgery and at 1, 3, and 6 months postsurgery. eGFR was calculated using the modification of diet in renal disease 2 equation modified for Japanese patients, as outlined by the Japanese Society of Nephrology: (eGFR = 1.94 × serum creatinine mg/dL^−1.094 × age × 0.739 [if female]).¹⁵

Surgery

All three surgical modalities were performed by five experienced surgeons. Surgical modality was decided mainly by the timing of surgery; LPN was adopted from 2000, and RAPN was adopted from 2013. After adoption of RAPN, most LPNs have been replaced by RAPN. OPN was chosen if LPN or RAPN was considered inappropriate in the clinical conference. The case volumes for each surgeon (A, B, C, D, and E) are listed in Table 1. Both LPN and RAPN were performed transperitoneally in most cases, and the retroperitoneal approach was used in posteriorly located tumors. OPN, LPN, and RAPN were performed using previously described techniques.^4,16

–19 En bloc renal hilar clamping was performed in LPN and OPN, and clamping of the main renal artery without renal veins was performed in most cases in RAPN. Moreover, 5 minutes of ice slush cooling was employed before tumor resection in OPN, which was included in the cold ischemia time. Renal tumors were resected with 2–5 mm of the renal parenchyma margin. A reconstruction inner suture was placed first; renorrhaphy was then performed after early unclamping in LPN and RAPN. Meanwhile, only an inner suture was placed, and the TachoSil tissue-sealing sheet (CSL Behring Japan, Tokyo, Japan) was sealed on the resection bed in OPN. Postoperative complication was defined by Clavien–Dindo classification.²⁰ Urine leakage was defined as persistent drain output >48 hours after PN with a chemical composition consistent with urine.²¹

Table 1.

Case Volumes of the Three Surgical Modalities for the Five Surgeons

	OPN	RAPN	LPN
A	82	5	19
B	43	114	7
C	33	116	20
D	137	388	25
E	292	336	144

LPN = laparoscopic partial nephrectomy; OPN = open partial nephrectomy; RAPN = robot-assisted partial nephrectomy.

Statistical analyses

Multivariate regression analysis assessed the association between the >25% decrease in eGFR within 1 month after surgery and potential preoperative factors (surgical procedure [LPN, OPN, or RAPN], preoperative eGFR, presence of proteinuria, hypertension [HT], diabetes mellitus [DM], solitary kidney, age, sex, Eastern Cooperative Oncology Group performance status [ECOG PS], RENAL nephrometry score, and tumor size) and surgeon-associated factors.²² To compare the renal functional outcome and minimize selection bias between the two groups (OPN vs RAPN and LPN vs RAPN), patient variables, including age, sex, body mass index (BMI), presence of HT, DM, proteinuria, and solitary kidney, American Society of Anesthesiologists Classification score, ECOG PS, tumor size, RENAL score, and preoperative eGFR, and surgeon-associated factors were adjusted using 1:1 propensity score matching. Propensity scores were calculated using a multivariate regression model. A caliper width of one-fifth of 1 standard deviation of the logit of the propensity score was used.²³ All statistical analyses were calculated with JMP Pro 14 (SAS Institute, Cary, NC). Student's t test was used for continuous variables, χ ² test for unordered categorical variables, and Mann–Whitney U test for ordinal categorical variables. A p value <0.05 was considered statistically significant.

Results

Patient and tumor background

The characteristics of 1762 patients (OPN: 588, LPN: 215, and RAPN: 959) before matching are presented in Table 2. Preoperative eGFR was significantly higher in LPN than in other surgical procedures (p = 0.0002). Among the surgical modalities, tumor size was significantly greater in OPN and less in LPN (p < 0.0001), and tumor complexity was significantly higher in OPN and lower in LPN (p < 0.0001). Furthermore, the rate of solitary kidney was significantly higher in OPN than in other surgical procedures (p < 0.0001).

Table 2.

Patient and Tumor Characteristics Before Matching

	Before matching
	OPN (N = 588)	RAPN (N = 959)	LPN (n = 215)	p Value
Age (years)	58 (±13)	57 (±13)	57 (±12)	0.59
Male (%)	411 (70)	682 (71.1)	148 (68.8)	0.75
Body mass index, kg/m²	23.9 (±3.8)	24.1 (±3.9)	24.1 (±4.0)	0.46
ECOG PS (0/1)	577 (98.1)/11 (1.9)	949 (99.0)/10 (1.0)	214 (99.6)/1 (0.4)	0.18
ASA score (1/2/3)	135 (23.0)/400 (68.0)/53 (9.0)	224 (23.3)/666 (69.5)/69 (7.2)	55 (25.6)/150 (69.8)/10 (4.7)	0.28
Presence of HT (%)	261 (44.4)	373 (38.9)	77 (35.8)	0.035
Presence of DM (%)	91 (15.5)	151 (15.7)	34 (15.8)	0.98
Presence of proteinuria (%)	73 (12.4)	91 (9.4)	9 (4.2)	0.0010
Preoperative eGFR, mL/min/1.73 m²	66 (±20)	67 (±18)	72 (±15)	0.0002
Preoperative chronic kidney disease
G1	58 (9.9)	81 (8.5)	25 (11.6)	<0.0001
G2	314 (53.4)	575 (60.0)	153 (71.2)
G3a	129 (21.9)	206 (21.5)	28 (13.0)
G3b	73 (12.4)	72 (7.5)	7 (3.3)
G4	14 (2.4)	23 (2.4)	2 (0.9)
G5	0 (0)	2 (0.2)	0 (0)
Tumor size (cm)	3.7 (±1.5)	3.0 (±1.3)	2.1 (±0.8)	<0.0001
RENAL score	8.2 (±1.8)	7.3 (±1.9)	6.1 (±1.5)	<0.0001
High	165 (28.1)	133 (13.9)	4 (1.9)	<0.0001
Intermediate	316 (53.7)	471 (49.1)	78 (36.2)	<0.0001
Low	107 (18.2)	355 (37.0)	133 (61.9)	<0.0001
R (1/2/3)	362 (61.6)/213 (36.2)/13 (2.2)	742 (77.3)/213 (22.2)/4 (0.4)	212 (98.6)/3 (1.4)/0 (0)	<0.0001
E (1/2/3)	161 (27.4)/303 (51.5)/124 (21.1)	273 (28.5)/542 (56.5)/144 (15.0)	94 (43.7)/108 (50.2)/13 (6.1)	<0.0001
N (1/2/3)	78 (13.3)/75 (12.8)/435 (74.0)	341 (35.5)/129 (13.5)/489 (51.0)	135 (62.8)/46 (21.4)/34 (15.8)	<0.0001
A (a/x/p)	79 (13.4)/431 (73.3)/78 (13.3)	318 (33.2)/350 (36.5)/291 (30.3)	42 (19.5)/151 (70.2)/22 (10.2)	<0.0001
L (1/2/3)	157 (26.7)/126 (21.4)/305 (51.9)	347 (36.2)/252 (26.3)/360 (37.5)	99 (46.0)/38 (17.7)/78 (36.3)	<0.0001
H	71 (12.1)	76 (7.9)	2 (0.9)	<0.0001
Solitary kidney (%)	49 (8.3)	6 (0.6)	3 (1.4)	<0.0001

AKI = acute kidney injury; ASA = American Society of Anesthesiologists; DM = diabetes mellitus; ECOG PS = Eastern Cooperative Oncology Group performance status; eGFR = estimated glomerular filtration rate; HT = hypertension.

Analysis of factors influencing AKI

Postoperative AKI (>25% decrease in eGFR) occurred in 382 cases (OPN: 237, LPN: 51, and RAPN: 94). Meanwhile, the % decrease in eGFR by ≥50% occurred in 50 cases (OPN: 39, LPN: 4, and RAPN: 7). The multivariate logistic regression analysis showed that LPN (odds ratio [OR]: 4.52) and OPN (OR: 4.48) were significantly associated with the incidence of postoperative AKI (p < 0.0001). Furthermore, solitary kidney (OR: 9.07, p < 0.0001), tumor size (OR: 1.29, p < 0.0001), male sex (OR: 1.71, p = 0.0006), high RENAL score (OR: 1.90, p = 0.0019), and proteinuria (OR: 1.69, p = 0.014) reached statistical significance. On the other hand, surgeon-associated factors did not reach statistical significance (Tables 3 and 4).

Table 3.

Incidence of Acute Kidney Injury in Each Surgical Modality

Acute kidney injury	OPN (N = 588)	LPN (N = 215)	RAPN (N = 959)	p Value
Risk	237 (40.3%)	51 (23.7%)	94 (9.8%)	<0.0001
Injury	39 (6.6%)	4 (1.8%)	7 (0.73%)	<0.0001
Failure	6 (1.0%)	2 (0.9%)	0	0.0018
Loss of kidney function	0	0	0	—
End-stage kidney disease	0	0	0	—

AKI was defined according to the RIFLE criteria: Risk, eGFR decrease >25%; Injury, eGFR decrease >50%; Failure, eGFR decrease >75%; Loss of kidney function, complete loss of kidney function for >4 weeks.

RIFLE = risk, injury, failure, loss, and end-stage kidney disease.

Table 4.

Predictive Factors Associated with Postoperative Acute Kidney Injury in the Multivariate Analysis

	OR	95% CI	p Value
Age (years)	1.00	0.99–1.02	0.19
Sex (male/female)	1.71	1.26–2.34	0.0006
Body mass index	1.00	0.96–1.04	0.86
ECOG PS
1	0.88	0.28–2.80	0.83
0	Reference	—	—
ASA score
3	0.92	0.49–1.70	0.79
2	1.08	0.75–1.55	0.65
1	Reference	—	—
Presence of HT	1.31	0.95–1.79	0.08
Presence of DM	0.93	0.57–1.20	0.32
Presence of proteinuria	1.69	1.10–2.58	0.014
Preoperative eGFR, mL/min/1.73 m²	1.01	1.00–1.01	0.016
Tumor size (cm)	1.29	1.17–1.43	<0.0001
RENAL score
High	1.90	1.26–2.85	0.0019
Intermediate	1.17	0.85–1.62	0.32
Low	Reference	—	—
Solitary kidney	9.07	4.59–17.9	<0.0001
Surgical procedure
LPN	4.52	2.93–6.97	<0.0001
OPN	4.48	3.31–6.07	<0.0001
RAPN	Reference	—	—
Surgeon
A	0.95	0.59–1.54	0.85
B	1.48	0.93–2.35	0.09
C	1.07	0.67–1.73	0.75
D	0.94	0.68–1.30	0.94
E	Reference	—	—

OR = odds ratio; CI = confidence interval.

Matched cohort analysis (OPN vs RAPN)

Of the 588 and 959 patients who underwent OPN and RAPN, respectively, 368 patients in each group were selected (Tables 5 and 6). After matching, the mean age was 58 years, mean tumor size was 3.4 cm, and preoperative eGFR was 66 mL/min/1.73 m². The mean RENAL score was 7.9 (Tables 5 and 6). Compared with OPN, RAPN had shorter operative (174 vs 199 minutes, p < 0.0001) and ischemia (20 vs 42 minutes, p < 0.0001) times, less estimated blood loss (84 vs 168 mL, p < 0.0001), shorter length of hospital stay (4.4 vs 6.3 days, p < 0.0001), lower incidence of AKI (13.6 vs 35.9%, p < 0.0001), and lower % decrease in eGFR after 3 months (−5.7% vs −8.8%, p = 0.021), although the % decrease in eGFR after 6 months did not differ significantly (−7.4% vs −8.7%, p = 0.33) (Tables 7 and 8). Subgroup analysis of preoperative eGFR ≥60 mL/min/1.73 m² and eGFR <60 mL/min/1.73 m² showed that the incidence of AKI was lower in RAPN in both subgroups. The % decrease in eGFR after 6 months was lower in RAPN than in OPN in the preoperative eGFR ≥60 mL/min/1.73 m² group (p = 0.010), although the difference disappeared after 12 months (p = 0.11). As for postoperative complications, the incidence of overall complications (20.7% vs 14.4%; p = 0.03) was significantly higher in OPN than in RAPN, although no significant differences were seen in urologic complications and major complications (Tables 9 and 10).

Table 5.

Patient and Tumor Characteristics After Matching Between Open Partial Nephrectomy and Robot-Assisted Partial Nephrectomy

	OPN (n = 368)	RAPN (n = 368)	p Value
Age (years)	58 (±13)	58 (±14)	0.99
Male (%)	270 (73.3)	273 (74.2)	0.8
Body mass index, kg/m²	24.0 (±3.8)	23.9 (±3.6)	0.77
ECOG PS (0/1)	361 (98.1)/7 (1.9)	362 (98.4)/6 (1.6)	0.77
ASA score (1/2/3)	85 (23.1)/252 (68.5)/31 (8.4)	81 (22.0)/255 (69.3)/32 (8.7)	0.93
Presence of HT (%)	160 (43.5)	168 (45.6)	0.60
Presence of DM (%)	54 (14.7)	61 (16.6)	0.54
Presence of proteinuria (%)	44 (12.0)	43 (11.7)	1.00
Preoperative eGFR, mL/min/1.73 m²	66 (±19)	66 (±19)	0.85
Preoperative chronic kidney disease
G1	34 (9.2)	33 (9.0)
G2	209 (56.8)	207 (56.3)
G3a	76 (20.6)	85 (23.1)
G3b	40 (10.9)	31 (8.4)
G4	9 (2.5)	11 (3.0)
G5	0 (0)	1 (0.3)	0.72
Tumor size (cm)	3.4 (±1.4)	3.4 (±1.3)	0.84
RENAL score	7.9 (±1.9)	7.9 (±1.7)	0.46
High	86 (23.4)	84 (22.8)	0.46
Intermediate	193 (52.5)	205 (55.7)	0.61
Low	89 (24.2)	79 (21.5)	0.61
R (1/2/3)	259 (70.4)/103 (28.0)/6 (1.6)	244 (66.3)/121 (32.9)/3 (0.8)	0.23
E (1/2/3)	101 (27.4)/192 (52.2)/75 (20.4)	86 (23.4)/211 (57.3)/71 (19.3)	0.33
N (1/2/3)	71 (19.3)/53 (14.4)/244 (66.3)	59 (16.0)/53 (14.4)/256 (69.6)	0.49
A (a/x/p)	63 (17.1)/241 (65.5)/64 (17.4)	65 (17.6)/242 (65.8)/61 (16.6)	0.94
L (1/2/3)	111 (30.2)/78 (21.2)/179 (48.6)	110 (29.9)/86 (23.4)/172 (46.7)	0.76
H	45 (12.2)	52 (14.1)	0.51
Solitary kidney (%)	9 (2.5)	5 (1.4)	0.42

Table 6.

Patient and Tumor Characteristics After Matching Between Laparoscopic Partial Nephrectomy and Robot-Assisted Partial Nephrectomy

	LPN (n = 147)	RAPN (n = 147)	p Value
Age (years)	56 (±12)	57 (±12)	0.62
Male (%)	104 (70.7)	105 (71.4)	0.89
Body mass index, kg/m²	24.2 (±3.9)	24.3 (±3.8)	0.94
ECOG PS (0/1)	146 (99.3)/1 (0.7)	147/0	0.23
ASA score (1/2/3)	42 (28.6)/99 (67.3)/6 (4.1)	44 (29.9)/95 (64.6)/8 (5.4)	0.81
Presence of HT (%)	52 (35.4)	54 (36.7)	0.80
Presence of DM (%)	20 (13.6)	21 (14.3)	0.86
Presence of proteinuria (%)	6 (4.1)	6 (4.1)	1.00
Preoperative eGFR, mL/min/1.73 m²	70 (±14)	70 (±15)	0.68
Preoperative chronic kidney disease
G1	11 (7.5)	15 (10.2)	0.12
G2	111 (75.5)	94 (63.9)
G3a	19 (12.9)	34 (23.1)
G3b	5 (3.4)	4 (2.7)
G4	1 (0.7)	0 (0)
G5	0 (0)	0 (0)
Tumor size (cm)	2.2 (±0.8)	2.2 (±0.9)	0.61
RENAL score	6.3 (±1.5)	6.1 (±1.6)	0.15
High	4 (2.7)	3 (2.0)	0.15
Intermediate	65 (44.2)	46 (31.3)	0.058
Low	78 (53.1)	98 (66.6)	0.058
R (1/2/3)	158 (98.1)/3 (1.9)/0 (0)	153 (95.0)/8 (5.0)/0 (0)	0.22
E (1/2/3)	52 (35.4)/82 (55.8)/13 (8.8)	55 (37.4)/84 (57.1)/8 (5.4)	0.52
N (1/2/3)	84 (57.1)/30 (20.4)/33 (22.5)	93 (63.3)/27 (18.4)/27 (18.4)	0.54
A (a/x/p)	34 (23.1)/94 (64.0)/19 (12.9)	39 (26.5)/90 (61.2)/18 (12.2)	0.79
L (1/2/3)	64 (43.5)/26 (17.7)/57 (38.8)	70 (47.6)/31 (21.1)/46 (31.3)	0.57
H	2 (1.4)	2 (1.4)	1.00
Solitary kidney (%)	1 (0.7)	2 (1.4)	0.55

Table 7.

Comparison of Perioperative Outcomes: Open Partial Nephrectomy vs Robot-Assisted Partial Nephrectomy and Laparoscopic Partial Nephrectomy vs Robot-Assisted Partial Nephrectomy

	OPN (n = 411)	RAPN (n = 411)	p Value	LPN (n = 147)	RAPN (n = 147)	p Value
Operative time (minutes)	199 (±56)	174 (±64)	<0.0001	204 (±52)	156 (±41)	<0.0001
Estimated blood loss (mL)	168 (±266)	84 (±165)	<0.0001	68 (±185)	44 (±72)	0.15
Ischemia time (minutes)	42 (±20)	20 (±9)	<0.0001	30 (±16)	15 (±6)	<0.0001
Mannitol use (%)	42 (11.4)	0	<0.0001	27 (18.4)	0	<0.0001
Length of hospital stay (days)	6.3 (±3.2)	4.4 (±1.8)	<0.0001	6.2 (4.5)	4.2 (±2.1)	<0.0001
% Decrease in eGFR	−19.3 (±19.0)	−9.4 (±15.4)	<0.0001	−13.1 (±18.9)	−–3.1 (±13.8)	<0.0001
Postoperative AKI
% Decrease >25% (%)	132 (35.9)	50 (13.6)	<0.0001	34 (23.1)	6 (4.1)	<0.0001
% Decrease >50% (%)	15 (4.1)	3 (0.8)	0.0069	3 (2.0)	1 (0.7)	0.30
% Decrease in eGFR
1 Month	−9.9 (±15.3)	−7.1 (±13.3)	0.030	−10.2 (±14.2)	−3.3 (±11.6)	<0.0001
3 Months	−8.8 (±18.7)	−5.7 (±11.7)	0.021	−8.5 (±10.6)	−2.9 (±10.3)	<0.0001
6 Months	−8.7 (±18.6)	−7.4 (±11.8)	0.33	−9.2 (±11.2)	−3.1 (±9.9)	<0.0001

Table 8.

Comparison of the Incidence of Acute Kidney Injury: Subgroup Analysis of the Preoperative Estimated Glomerular Filtration Rate ≥60 and Estimated Glomerular Filtration Rate <60

Preoperative eGFR	OPN	RAPN	p Value	LPN	RAPN	p Value
Preoperative eGFR ≥60, N (%)	243	240		122	109
Solitary kidney, N (%)	2 (0.8)	0	0.49	0	0	—
% Decrease >25%, N (%)	88 (36.2)	25 (10.4)	<0.0001	27 (22.1)	4 (3.7)	<0.0001
% Decrease in eGFR (nadir)	−20.4 (±15.6)	−9.1 (±13.7)	<0.0001	−13.7 (±17.1)	−4.3 (±13.6)	<0.0001
% Decrease in eGFR (6 months)	−11.0 (±12.4)	−7.8 (±9.6)	0.010	−9.9 (±10.5)	−5.1 (±10.3)	0.0019
Preoperative eGFR <60, N (%)	125	128		25	38
Solitary kidney, N (%)	7 (5.6)	5 (3.9)	0.49	2 (8.0)	1 (2.6)	0.55
% Decrease >25% N (%)	44 (35.2)	25 (19.5)	0.0071	7 (28.0)	2 (5.3)	0.023
% Decrease in eGFR (nadir)	−17.0 (±24.1)	−10.0 (±18.3)	0.0098	−10.2 (±26.7)	0.1 (±13.9)	0.049
% Decrease in eGFR (6 months)	−3.8 (±26.9)	−6.5 (±14.2)	0.46	−7.4 (±14.2)	−0.2 (±12.4)	0.052

Table 9.

Comparison of Perioperative Complications (Open Partial Nephrectomy vs Robot-Assisted Partial Nephrectomy)

Perioperative complications	OPN (n = 368)	RAPN (n = 368)	p Value
Overall complications (Clavien 1–5)	76 (20.7%)	53 (14.4%)	0.03
Major complications (Clavien 3–5)	22 (5.9%)	19 (5.2%)	0.74
Urologic complications	58 (15.8%)	40 (10.9%)	0.06

Clavien grade	Any	3–5	Any	3–5
Hematoma	8 (2.2)	2 (0.3)	10 (2.7)	0
Hematuria	2 (0.5)	0	3 (0.8)	1 (0.24)
Asymptomatic pseudoaneurysm	22 (6.0)	13 (3.5)	24 (6.5)	15 (4.1)
Urine leak	25 (6.8)	8 (2.2)	0	0
Ureteral obstruction	0	0	1 (0.3)	1 (0.3)
Urinary tract infection	1 (0.3)	0	0	0
Wound infection	6 (1.6)	0	3 (0.8)	0
Pulmonary thromboembolism	1 (0.3)	1 (0.3)	1 (0.3)	0
Cerebral infarction	0	0	1 (0.3)	1 (0.3)
Arterial stent thromboemboism	0	0	1 (0.3)	1 (0.3)
Arrythmia	0	0	1 (0.3)	1 (0.3)
Pneumothorax	1 (0.3)	1 (0.3)	0	0
Pleural effusion	2 (0.5)	0	0	0
Bowel injury	0	0	1 (0.3)	1 (0.3)
Ileus	3 (0.8)	1 (0.3)	0	0
Liver dysfunction	2 (0.5)	0	4 (1.1)	0

Table 10.

Comparison of Perioperative Complications (Laparoscopic Partial Nephrectomy vs Robot-Assisted Partial Nephrectomy)

Perioperative complications	LPN (n = 147)	RAPN (n = 147)	p Value
Overall complications (Clavien 1–5)	29 (19.7%)	20 (13.6%)	0.21
Major complications (Clavien 3–5)	17 (11.6%)	7 (4.8%)	0.03
Urologic complications	26 (17.7%)	15 (10.2%)	0.09

Clavien grade	Any	3–5	Any	3–5
Hematoma	4 (2.7)	1 (0.7)	5 (3.4)	0
Hematuria	0	0	2 (1.4)	0
Asymptomatic pseudoaneurysm	19 (12.9)	13 (8.8)	12 (8.8)	7 (4.8)
Urine leak	3 (2.0)	3 (1.4)	0	0
Wound infection	0	0	1 (0.7)	0
Arrythmia	0	0	2 (1.4)	0
Angina pectoris	0	0	1 (0.7)	0
Pneumothorax	1 (0.7)	1 (0.7)	0	0
Pneumonia	1 (0.7)	0	0	0
Bowel injury	1 (0.7)	1 (0.7)	1 (0.7)	1 (0.7)
Ileus	1 (0.7)	0	0	0
Liver dysfunction	1 (0.7)	0	0	0

Matched cohort analysis (LPN vs RAPN)

Of the 215 and 959 patients who underwent LPN and RAPN, respectively, 147 in each group were selected. After matching, the mean age was 56 (LPN) and 57 (RAPN) years, mean tumor size was 2.2 cm, and preoperative eGFR was 70 mL/min/1.73 m². The mean RENAL score was 6.3 (LPN) and 6.1 (RAPN). Compared with LPN, RAPN had shorter operative (156 vs 204 minutes, p < 0.0001) and warm ischemia (15 vs 30 minutes, p < 0.0001) times, shorter length of hospital stay (4.2 vs 6.2 days, p < 0.0001), lower incidence of AKI (4.1% vs 23.1%, p < 0.0001), and lower % decrease in eGFR after 3 months (−2.9% vs −8.5%, p < 0.0001) and 6 months (−3.1% vs −9.2%, p < 0.0001) (Tables 7 and 8). As for postoperative complications, the incidence of major complications (Clavien–Dindo grade ≥3) was higher in LPN than in RAPN (p = 0.03), although no significant differences were seen in urologic and overall complications.

Discussion

OPN has been used for localized renal tumors since 1990. With the prevalence of minimally invasive surgery, LPN has been adopted since 2000 and RAPN since 2013 at our institutions. In this study, the surgical modality (OPN and LPN), solitary kidney, and large and highly complex tumors were associated with postoperative AKI as predictive factors. After propensity score matching and comparison between the two matched cohorts (LPN vs RAPN and OPN vs RAPN), LPN was likely to have longer ischemia time and more frequently resulted in AKI, compared with RAPN, and affected long-term renal function more. Meanwhile, OPN was likely to have a long ischemia time and resulted in AKI more frequently compared with RAPN, but that did not affect long-term renal function. Our data showed that RAPN was less likely to cause AKI compared with other surgical modalities, although AKI after OPN could recover in the long term.

Predictive factors for postoperative renal functional decrease after RAPN have been published in several reports. Martini and colleagues²⁴ developed a nomogram to predict significant eGFR reduction 3–15 months after RAPN, using data of 999 patients from a multi-institutional database. They included older age, high RENAL score, high preoperative eGFR, and postoperative AKI in their nomogram. They showed that the occurrence of AKI after RAPN affects long-term renal function.²⁴ Furthermore, they developed another nomogram to estimate the probability of postoperative AKI after RAPN. They included older age, male sex, high BMI, diabetes, and high preoperative eGFR in their nomogram.²⁵ Our study showed similar results, that is, male sex, proteinuria, high preoperative eGFR, high RENAL score, solitary kidney, and OPN or LPN were predictive of AKI. Most of them are self-limiting conditions; therefore, we sought to identify the differences in surgical modalities selected for patients.

Difference in ischemia time

Compared with LPN, RAPN had shorter operative and warm ischemia times and shorter hospital stay. AKI incidence was lower, and % decreases in eGFR after 1, 3, and 6 months were lower in RAPN than in LPN. Given the long learning curve in LPN,²⁶ the indication for LPN at our institutions was limited to small exophytic tumors. Therefore, the matched cohorts had small tumor sizes (average 2.2 cm) and low RENAL scores (average 6.1–6.3). The difference in the incidence of AKI and subsequent long-term renal function may be caused by long warm ischemic time and large parenchymal loss in LPN, as analyzed in our previous study.⁴ Compared with OPN, RAPN has shorter operative time, lower estimated blood loss, and shorter hospital stay, which may be caused by the minimal invasiveness of RAPN. Furthermore, RAPN has shorter ischemia time and lower AKI incidence, which may be explained by the ice slush cooling performed for 5 minutes before tumor resection to obtain adequate core cooling in OPN and the surgical invasiveness of open surgery.

Difference in clamping technique

Differences in the renal clamping technique may affect the incidence of AKI: en bloc renal hilar clamping with cold ischemia in OPN, en bloc renal hilar clamping with warm ischemia in LPN, and clamping of the main renal artery without the renal vein in RAPN. Whether to clamp only the renal artery or renal hilum (both artery and vein) for renal functional preservation remains controversial. In porcine model studies, Schuler et al. reported no significant difference in the renal pO₂ arterial level or use of arteriovenous clamping, although the renal pO₂ level after cold ischemia slowly returned to baseline compared with warm ischemia.²⁷ Meanwhile, Orvieto and colleagues reported that clamping only the renal artery resulted in lower serum creatinine increase within 3 days after surgery compared with arteriovenous clamping.²⁸ At our institutions, in OPN, en bloc renal hilar clamping has been employed for various reasons: vasospasm from renal artery-separating procedures, arterial endothelial injury from direct renal artery clamping, or difficulty in maintaining a bloodless field during tumor resection and reconstruction.^29,30 In the era of minimally invasive surgery, Gill and colleagues reported en bloc renal hilar clamping with warm ischemia in LPN, and this technique resulted in renal functional outcome equivalent to OPN.¹⁷ As for RAPN, the renal artery-separating procedure becomes easier and less invasive; furthermore, due to the pneumoperitoneal pressure, a bloodless field can be maintained with only renal artery clamping. Blood supply of backflow from the renal vein during tumor resection and reconstruction may help maintain the renal pO₂ level and may reduce the risk of postoperative AKI.

Effects on long-term renal function

Interestingly, in this study, no significant difference was identified in the % decrease in eGFR 6 months after surgery, although the % decrease in eGFR at 1 and 3 months differed significantly between RAPN and OPN. Subgroup analysis of preoperative eGFR <60 mL/min/1.73 m² and eGFR ≥60 mL/min/1.73 m² showed similar results. RAPN had lower % decrease in eGFR after 6 months in the group with preoperative eGFR ≥60 mL/min/1.73 m², but the difference disappeared after 12 months. Although longer ischemia time might result in AKI in OPN, renal function could recover in the long term if the cold ischemia technique is used. Zabell and colleagues⁶ analyzed the association of postoperative AKI and long-term renal function in a solitary kidney cohort. Their multivariable analysis showed that AKI did not affect long-term functional change; meanwhile, warm ischemia was associated with long-term functional decline. The possible mechanism for the deleterious effect of warm ischemia has been explained: nephrons affected by warm ischemia are more fragile and less able to recover than those affected by cold ischemia.⁶ With regard to perioperative complications, the overall complication rate was lower in RAPN than in OPN, which was caused by the higher incidence of urine leak in OPN. In OPN, renorrhaphy was omitted, with only an inner suture with TachoSil tissue-sealing sheet (CSL Behring Japan) compression; meanwhile, renorrhaphy was added after inner suture placement in RAPN. The lack of an outer suture may affect the incidence of urine leakage.¹⁹ From our study, by comparing each surgical modality, the incidence of AKI was lower in RAPN than in OPN in small and low complexity renal tumors. Most AKI cases that occurred in patients with normal baseline renal function will recover conservatively. Our study suggests that patients with chronic kidney disease or a solitary kidney who are susceptible to AKI, which may lead to temporary hemodialysis or cardiovascular disease, can benefit from RAPN. Meanwhile, OPN will be useful in cases with an expectedly long ischemia time: multiple renal tumors or large and highly complex tumors.

Limitations

This study had several limitations, such as its retrospective design, collection of data from two high-volume academic and teaching hospitals, and study population comprising tertiary care patients. All the unmeasured variables and confounders could not be excluded in the study of real-world clinical practice. The surgical modality was mainly selected based on the timing of surgery. Moreover, surgeon-associated bias could not be excluded, although we included five surgeons in the propensity score matching. Technical details were different in each surgical modality. The da Vinci surgical system enabled us to easily clamp only the renal artery. Although cold ischemia is easy in OPN, it is difficult in LPN and RAPN. OPN enabled the nonrenorrhaphy technique (only an inner suture) by manual compression of the TachoSil tissue-sealing sheet for hemostasis. Although the solitary kidney model may be optimum for this study according to previous reports, our study population had renal tumors with normal contralateral kidneys.^5,6 The function of the operated kidneys was not evaluated with renal scintigrams before or after surgery, but with global kidney function (eGFR). The RIFLE criteria may not be appropriate for the definition of AKI after PN because we did not calculate resected parenchymal volume loss to estimate postoperative new baseline renal function.⁵ However, as the aim of this study was to compare the incidence of AKI among surgical modalities, the analysis of % decrease in eGFR will be acceptable for this outcome. As strengths of this study, patient and tumor variables were adjusted using propensity score matching and two matched cohorts, RAPN vs LPN and RAPN vs OPN, were formed to decrease selection bias, although the RENAL score did not match completely.

Conclusions

The incidence of AKI was lower in RAPN than in OPN in small and low complexity renal tumors. Meanwhile, the long-term renal outcome did not differ between RAPN and OPN, indicating that OPN is still a valuable treatment option without compensating renal function. RAPN may be useful for patients with preoperative chronic kidney disease who are at a high risk of postoperative AKI that may lead to the requirement for emergent hemodialysis. Our study will help physicians determine surgical modalities during preoperative patient counseling, considering the tumor size, tumor complexity, patient comorbidities, and surgeon experience.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Abbreviations Used

References

European Association of Urology. European Association of Urology Guidelines. 2018 Edition. Vol. Presented at the EAU Annual Congress Copenhagen 2018. Arnhem, The Netherlands: European Association of Urology Guidelines Office, 2018.

Huang

, Elkin

, Levey

, et al. Partial nephrectomy versus radical nephrectomy in patients with small renal tumors—is there a difference in mortality and cardiovascular outcomes?. J Urol, 2009; 181:55–61; discussion 61–66.

Weight

, Lieser

, Larson

, et al. Partial nephrectomy is associated with improved overall survival compared to radical nephrectomy in patients with unanticipated benign renal tumours. Eur Urol, 2010; 58:293–298.

Tachibana

, Takagi

, Kondo

, et al. Robot-assisted laparoscopic partial nephrectomy versus laparoscopic partial nephrectomy: A propensity score-matched comparative analysis of surgical outcomes and preserved renal parenchymal volume. Int J Urol, 2018; 25:359–364.

Zhang

, Zhao

, Dong

, et al. Acute kidney injury after partial nephrectomy: Role of parenchymal mass reduction and ischemia and impact on subsequent functional recovery. Eur Urol, 2016; 69:745–752.

Zabell

, Isharwal

, Dong

, et al. Acute kidney injury after partial nephrectomy of solitary kidneys: Impact on long-term stability of renal function. J Urol, 2018; 200:1295–1301.

Masson-Lecomte

, Bensalah

, Seringe

, et al. A prospective comparison of surgical and pathological outcomes obtained after robot-assisted or pure laparoscopic partial nephrectomy in moderate to complex renal tumours: Results from a French multicentre collaborative study. BJU Int, 2013; 111:256–263.

, Li

, Song

, et al. Propensity-score matched analysis comparing robot-assisted with laparoscopic partial nephrectomy. BJU Int, 2015; 115:437–445.

Choi

, You

, Kim

, et al. Comparison of perioperative outcomes between robotic and laparoscopic partial nephrectomy: A systematic review and meta-analysis. Eur Urol, 2015; 67:891–901.

10.

Shiroki

, Fukami

, Fukaya

, et al. Robot-assisted partial nephrectomy: Superiority over laparoscopic partial nephrectomy. Int J Urol, 2015; 23:122–131.

11.

Woldu

, Thoreson

, Okhunov

, et al. Comparison of renal parenchymal volume preservation between partial nephrectomy, cryoablation, and radiofrequency ablation. J Endourol, 2015; 29:948–955.

12.

Zargar

, Akca

, Ramirez

, et al. The impact of extended warm ischemia time on late renal function after robotic partial nephrectomy. J Endourol, 2014; 29:444–448.

13.

Dagenais

, Maurice

, Mouracade

, et al. The synergistic influence of ischemic time and surgical precision on acute kidney injury after robotic partial nephrectomy. Urology, 2017; 107:132–137.

14.

Bellomo

, Ronco

, Kellum

, et al. Acute renal failure-definition, outcome measures, animal models, fluid therapy and information technology needs: The Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care, 2004; 8:R204–R212.

15.

Matsuo

, Imai

, Horio

, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis, 2009; 53:982–992.

16.

Kondo

, Takagi

, Morita

, et al. Early unclamping might reduce the risk of renal artery pseudoaneurysm after robot-assisted laparoscopic partial nephrectomy. Int J Urol, 2015; 22:1096–1102.

17.

Gill

, Desai

, Kaouk

, et al. Laparoscopic partial nephrectomy for renal tumor: Duplicating open surgical techniques. J Urol, 2002; 167:469–477; discussion 475–476.

18.

Omae

, Kondo

, Takagi

, et al. Renal sinus exposure as an independent factor predicting asymptomatic unruptured pseudoaneurysm formation detected in the early postoperative period after minimally invasive partial nephrectomy. Int J Urol, 2015; 22:356–361.

19.

Tachibana

, Takagi

, Kondo

, et al. Comparison of perioperative outcomes with or without renorrhaphy during open partial nephrectomy: A propensity score-matched analysis. Int Braz J Urol, 2018; 44:467–474.

20.

Dindo

, Demartines

, Clavien

. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg, 2004; 240:205–213.

21.

Meeks

, Zhao

, Navai

, et al. Risk factors and management of urine leaks after partial nephrectomy. J Urol, 2008; 180:2375–2378.

22.

Hung

, Cai

, Simmons

, et al. “Trifecta” in partial nephrectomy. J Urol, 2013; 189:36–42.

23.

Faries

, Leon

, Haro

, et al. Analysis of Observational Health Care Data Using SAS. SAS Institute Inc., Cary, North Carolina, 2010.

24.

Martini

, Cumarasamy

, Beksac

, et al. A nomogram to predict significant estimated glomerular filtration rate reduction after robotic partial nephrectomy. Eur Urol, 2018; 74:833–839.

25.

Martini

, Sfakianos

, Paulucci

, et al. Predicting acute kidney injury after robot-assisted partial nephrectomy: Implications for patient selection and postoperative management. Urol Oncol, 2019; 37:445–451.

26.

Kern

, Lustik

, McMann

, et al. Comparison of outcomes after minimally invasive versus open partial nephrectomy with respect to trainee involvement utilizing the American College of Surgeons National Surgical Quality Improvement Program. J Endourol, 2014; 28:40–47.

27.

Schuler

, Perks

, Fazio

, et al. Impact of arterial and arteriovenous renal clamping with and without intrarenal cooling on renal oxygenation and temperature in a porcine model. J Endourol, 2008; 22:2367–2372.

28.

Orvieto

, Zorn

, Mendiola

, et al. Recovery of renal function after complete renal hilar versus artery alone clamping during open and laparoscopic surgery. J Urol, 2007; 177:2371–2374.

29.

Manship

, Moore

, Bynoe

, et al. Differential endothelial injury caused by vascular clamps and vessel loops. II. Atherosclerotic vessels. Am Surg, 1985; 51:401–406.

30.

Simmons

, Schreiber

, Gill

. Surgical renal ischemia: A contemporary overview. J Urol, 2008; 180:19–30.