Background
Scholars agree that the partial nephrectomy for small renal mass is advantageous over radical nephrectomy in terms of renal function [
1,
2]. The conventional partial nephrectomy technique includes the clamping of the renal artery (on-PN); this method allows tumor resection and renal reconstruction in a relatively bloodless field [
3‐
5]. However, occluding the renal artery places the remaining nephrons at risk of ischemia−reperfusion injury and mitigates the renoprotective purpose of surgery [
3,
6]. Shorter ischemia intervals have been correlated with better renal functional preservation [
7,
8]. Off-clamp partial nephrectomy (off-PN) has been a common strategy to avoid ischemia in small renal tumor. Although a consensus has been reached on off-PN risking more blood loss, the impact of the two methods on the change of postoperative short- and long-term renal function remains unclear [
1,
9]. Therefore, we systematically searched and analyzed the clinical studies comparing off-PN with on-PN for small renal masses published until January 2018 to assess the methods’ impact on short- and long-term renal function.
Methods
The literature search methods, inclusion and exclusion criteria, outcome measures, and statistical analysis methods were well defined in a prospective protocol.
Literature-search strategy
A literature search was performed in January 2018 with no restriction to region or language. The primary sources were the electronic databases of PubMed, EMBASE, and the Cochrane Library. The following terms and their combinations were searched as follows: (“partial nephrectomy” OR “nephron sparing surgery” [Title/Abstract]) and (“clamp*” OR “ischemia” [Title/Abstract]). Our computer search was supplemented with manual searches of reference lists of all retrieved review articles. When multiple studies were reported by the same institution and/or authors, the most complete report was included in our analysis.
Inclusion and exclusion criteria
All retrospective or prospective comparative studies (cohort or case−control studies) containing a comparison of off-PN and on-PN with or without a third group, such as cold-ischemia partial nephrectomy, and those providing available data to assess postoperative renal function, were included. Animal experimental studies, editorials, letters to the editor, review articles, case reports, conference abstracts, studies without available data about postoperative renal function, and non-comparative studies were excluded.
Data extraction and outcome measures
Two of the authors extracted and summarized data from the included studies independently. Any disagreement was resolved by mutual discussion with another two adjudicating senior authors.
The primary outcomes were short-term change of estimated glomerular filtration rate (eGFR), short-term change of Cr level, long-term change of eGFR, long-term % change of eGFR, and long-term change of Cr level.
The secondary outcomes were postoperative long-term eGFR, short-term eGFR, short-term Cr level, postoperative increase in CKD (stage≥3), and postoperative acute renal failure (ARF).
Quality assessment and statistical analysis
The modified Newcastle−Ottawa scale [
10,
11] was used to assess the methodological quality of nonrandomized studies comprising three factors: patient selection, comparability of study groups, and assessment of outcome. Every study was scored from 0 to 9.
The level of evidence was assessed on the basis of the criteria enacted by the Center for Evidence-Based Medicine in Oxford used to rate the included studies [
12].
Review Manager 5.3 (Cochrane Collaboration, Oxford, UK) was used to perform all meta-analysis. The odds ratio (OR) with 95% confidence interval (CI) was used to compare the dichotomous variables consisting of ARF and postoperative increase in chronic kidney disease (CKD; stage≥3), and the weighted mean difference (WMD) with 95%CI was used to compare the remaining continuous variables. The corresponding authors were contacted when the data were missing or incomplete. The technique summarized by Hozo et al. was used to convert medians to means [
13].
Statistical heterogeneity was considered significant when the Cochrane Q test
p value was < 0.10. The standard heterogeneity test, I
2 statistic, was used to assess the consistency of the effect sizes. The fixed-effects model was used when no significant heterogeneity exists between the studies; otherwise, the random-effects model is used [
14].
Subgroup analysis was performed to verify the impact of two surgical procedures and to assess the efficacy of different studies in more homogeneous subsets in accordance with the sample size in the studies.
Sensitivity analysis was performed by repeating the primary analysis without including the highest scored study or studies when some papers achieved the highest scores at the same time.
STATA SE 12.0 was then utilized to evaluate potential publication bias, which was screened on funnel plots and assessed statistically using the Begg’s test and Egger’s test. The tests were two sided, and the p values of < 0.05 were considered significant statistically.
Discussion
Nephron-sparing surgery has been the recommended treatment of clinical T1a renal masses and favored over radical nephrectomy in patients with T1b tumors when technically feasible [
39]. Every minute counts to preserve renal function when the renal hilum is clamped during partial nephrectomy [
7,
40,
41]. For minimizing ischemic injury as much as possible to preserve the functional renal nephron, avoiding ischemia altogether by performing surgery entirely off-clamp is a good strategy. Many studies concluded that partial nephrectomy for small renal masses can be performed without hilar clamping, although considered as a procedure of potentially increased risk of augmented blood loss [
2,
3,
5,
9,
15‐
38]. However, the impact of off-PN on short- and long-term renal functional residual capacity compared with conventional on-PN remains unclear [
27]. Thus, we reviewed the published studies and conducted a standard meta-analysis to evaluate the impact of off-PN on short- and long-term renal function compared with conventional on-PN.
Within the postoperative 3 months, the short-term decrease in eGFR was found in both groups in almost all the eligible studies [
19,
24,
26,
28,
32,
34,
35,
38], except for Taweemonkongsap et al. [
9] and Kaczmarek et al. [
27]. The postoperative short-term eGFR of the off-PN group reported by Taweemonkongsap et al. [
9] and Kaczmarek et al. [
27] achieved an increase compared with preoperative eGFR (mean: 2.225 ml/min/1.73 m
2 and 1.4 ml/min/1.73 m
2, respectively). In the two studies, preoperative eGFR records in the off-PN group were higher, and all surgeries were robot assisted, which promoted an enhanced and hastened postoperative renal functional recovery. In all included studies, the short-term decrease of eGFR in the on-PN group was more than that in the off-PN group.
Although no significant difference was found for the postoperative short-term eGFR between two groups in the two studies [
24,
35], the pooled analysis of postoperative short-term eGFR showed a significantly higher eGFR in the off-PN group (WMD: 9.72; 95%CI: 6.13 to 13.30;
p < 0.00001). This result was consistent with a significantly higher decrease of the pooled short-term eGFR in the on-PN group.
The pooled analysis for short-term change of Cr level indicated a significantly higher Cr level increase in the on-PN group compared with that in the off-PN group. While most of the included studies showed a higher increase in the on-PN group [
15,
17,
20,
23,
24,
26,
28,
34], the remaining study reported by Kane et al. [
16] showed a higher increase, which was probably driven more by a 38% larger tumor size than by the effect of renal artery occlusion in the off-PN group than in the on-PN group. All included studies showed a higher postoperative short-term Cr level in the on-PN group than in the off-PN group in accordance with a pooled meta-analysis for the postoperative short-term Cr level; this result also indicates the better renal functional outcome of off-PN.
After 6 months or more of surgery, although long-term decrease of eGFR and percent eGFR in both groups was found in the included studies, the pooled meta-analysis for postoperative long-term eGFR change and percent eGFR change revealed a significantly greater decrease in the on-PN group than in the off-PN group. Between-study heterogeneity was significant for long-term eGFR change, the random-effects model was then utilized to reduce the effect of heterogeneity, whereas the long-term percent eGFR change was only the opposite. Subgroup analysis for both long-term eGFR change and percent eGFR change showed that the included large sample studies [
22,
25,
30‐
33] and small sample studies [
9,
19,
29,
36,
38] was pooled without significant between-study heterogeneity. The same conclusions were drawn in a large-sample subgroup with original pooled analysis, whereas no significant difference was found between two groups in the small sample subgroup. Sample size was considered a reason for heterogeneity, and a large sample subgroup is believed to be close to the truth. Sensitivity analysis was performed by exclusion of the highest scored study or studies in the modified Newcastle−Ottawa Scale. No change in the significance of the outcome was noted.
No significant difference was found between two groups for postoperative long-term Cr change. The random-effects model was used to pool the included studies owing to the significant difference of between-study heterogeneity. We attributed this result to the limited number of included studies. Sensitivity analysis with exclusion of the highest scored study [
38] found a significant difference between the two groups about long-term Cr change and a very low degree of between-study heterogeneity. Additional studies are needed to confirm the conclusion.
The pooled data of postoperative increase in CKD (stage≥3) suggests the lack of significant difference between the two groups, and the degree of between-study heterogeneity was moderate. In fact, eGFR after on-PN may not decrease to 60 ml/min/1.73 m2 or lower even when renal ischemia−reperfusion injury truly exists.
The ARF rate is significantly lower for the off-PN group than for the on-PN group. This result may be explained by the injury caused by renal ischemia during operation.
To assess any impact of the highest scored study or studies on the effect on primary outcomes, we performed a sensitivity analysis with exclusion of the highest scored study or studies. Given the moderate degree of between-study heterogeneity, all results performed with the fixed-effects model were similar to those of the original analysis except for the long-term Cr change, which was significantly increased in the on-PN group than in the off-PN group (WMD: − 0.09; 95%CI: − 0.13 to − 0.04; p = 0.0002).
Limitations and strength
The primary limitation of this systematic review and meta-analysis was that no RCTs were included for evaluation and subsequent analysis; hence, sufficient data are difficult to acquire for meaningful results. Moreover, the studies that provide data on the change in postoperative long-term Cr level were exceedingly few to offer a more convincing result than currently attained. In addition, the operations were performed by surgeons with different levels of surgical expertise and different choices of surgical approaches. Finally, patient allocation and treatment assignment were usually based on the physician’s attitude instead of randomized allocation; this aspect led to a significant selection bias.
Although a small number of papers have compared the two surgical procedures, no paper has discussed their long-term impact on renal function; this topic is particularly important to effectively choose the proper treatment when possible. An increasing number of T1-T2a peripheral renal tumors have been detected at a young age, and an improved outcome is meaningful. This meta-analysis was conducted at an appropriate time. A sufficient number of studies have been accumulated for inspection by meta-analytical methods. Studies were identified using multiple strategies; the methodological quality of the studies was evaluated on the basis of strict inclusion and exclusion. Subgroup and sensitivity analyses were performed to analyze the source of heterogeneity. The MOOSE guidelines were used to report our systematic review. Publication bias was not significant.