Right Versus Left Laparoscopic Living-Donor Nephrectomy: A Meta-Analysis

Transcription

1 ARTICle Right Versus Left Laparoscopic Living-Donor Nephrectomy: A Meta-Analysis Kun Wang, Peijin Zhang, Xianlin Xu, Min Fan Abstract Objectives: We assessed the safety and efficacy of right versus left laparoscopic living-donor nephrectomy. Few clinical-controlled studies have compared the right and left side, and most trials have a small sample number and varied results. A metaanalysis of published trials was performed to determine the effects of the 2 different approaches. Materials and Methods: Major databases including Medline (PubMed), Embase, Ovid, and Cochrane were searched to identify studies comparing right and left laparoscopic living-donor nephrectomy (January 2000 to January 2014). Outcomes evaluated included operative time, warm ischemia time, operative blood loss, 1-year graft loss, donor intraoperative and postoperative complications, recipient postoperative complications, donor blood transfusion, conversion to open donor nephrectomy, length of donor hospital stay, and delayed graft function. Results: There were 15 studies included with 3073 patients (left, 2420 patients [78%]; right, 653 patients [22%]). The right group had shorter operative time (weighted mean difference, min; 95% confidence interval, to min; P =.005) and lower operative blood loss (weighted mean difference, ml; 95% confidence interval, to ml; P =.003) than the left group. There was a higher rate of overall donor intraoperative complications in the left group (odds ratio, 0.53; From the Department of Surgical Urology, The Third Affiliated Hospital of Soochow University, Jiangsu Changzhou, ; and Department of Public Health, Xuzhou Medical College, Xuzhou, China Acknowledgements: The authors have no conflicts of interest to declare. No funding was received for this study. Corresponding author: Xianlin Xu, Department of Surgical Urology, The Third Affiliated Hospital of Soochow University, Jiangsu Changzhou, , China Phone: Fax: xuxianlin2014@sina.com Experimental and Clinical Transplantation (2015) 3: % confidence interval, 0.31 to 0.92; P =.03). There were no differences between groups in hospital stay, delayed graft function, recipient 1-year graft loss, conversion to open donor nephrectomy, donor blood transfusion, and donor or recipient postoperative complications. Conclusions: Right and left laparoscopic living-donor nephrectomy were similar in the effect of surgery and postoperative graft function. When there are no differences in bilateral renal function, surgeons can transplant the right or left kidney. However, the longer renal vein of the left kidney could decrease operative difficulty, and we recommend using the left kidney in clinical practice. Key words: End-stage renal disease, Transplant Introduction Living-donor kidney transplant is a surgical treatment for end-stage renal failure. 1,2 The 2 operative methods for the donor kidney include open and laparoscopic living-donor nephrectomy. Laparoscopic living-donor nephrectomy was introduced in Since then, laparoscopic living-donor nephrectomy has been performed increasingly in many medical institutions due to the advantages of laparoscopic surgery. 4-7 The laparoscopic living-donor nephrectomy includes traditional laparoscopic, retroperitoneoscopic, and hand-assisted living-donor nephrectomy. 8 However, there is controversy about the intraoperative and postoperative effects of each side. Some authors reported that the left side had many intraoperative and postoperative advantages due to the longer renal vein. 9,10 Other studies revealed that laparoscopic procurement can be performed with donor and graft outcomes similar for right and left kidney Therefore, we believed that a meta-analysis from relevant published trials was required to evaluate the Copyright Başkent University 2015 Printed in Turkey. All Rights Reserved. DOI: /ect

2 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: effects of the 2 different approaches. Subgroup analyses (comparing right and left laparoscopic, retroperitoneoscopic, and hand-assisted nephrectomy) and sensitivity analyses (including high quality studies and only randomized controlled trials) were performed. Materials and Methods Study inclusion Medline (PubMed), Embase, Ovid, Cochrane, and Chinese Biomedical Literature databases were searched for studies performed between 2000 and 2014 that compared right and left laparoscopic living-donor nephrectomy. The following medical subject headings were used, alone or in combination: laparoscopic, nephrectomy, renal transplant, comparative study, live donor, left, and right. The function of related articles was applied to enlarge the search. The references of each included study were reviewed, and the last study date for the search was May To ensure that any relevant studies were not missed, many laparoscopic surgery scholars were consulted. This analysis only included comparative clinical full-text studies, and the final articles included in the study were agreed upon by all authors of this study. Data extraction The following data were independently extracted from each study by 2 coauthors (PZ and KW): year of publication, first author, characteristics of targeted population, research design, interventions, and outcomes of interest. To avoid missing related study data, the 2 coauthors attempted to contact study authors when information was lacking or unclear. Conflicts between investigators about the outcomes of interest were reviewed and agreements were reached on the final interpretation of the data. Inclusion criteria Studies included in the analysis fulfilled the following criteria: (1) comparison of right versus left laparoscopic living-donor nephrectomy in patients undergoing living-donor nephrectomy; (2) human studies, and donors studied had no differences in bilateral renal function; (3) inclusion of 3 of the following outcome measures: operative time, operative blood loss, warm ischemia time, donor intraoperative complications, hospital stay, delayed graft function, recipient 1-year graft loss, recipient creatinine level (mg/dl) at 7 days, 1 month, and 1 year after transplant, conversion to open donor nephrectomy, donor blood transfusion, donor postoperative complications, and recipient pos-toperative complications; (4) documented operative technique as laparoscopic, hand-assisted, or retroperitoneoscopic; (5) the articles were published in English or Chinese in open access journals; and (6) when multiple trials were from the same authors and/or institute, only the latest publication or publication of highest quality was included in the study. exclusion criteria Trials were excluded for the following reasons: (1) case reports, reviews, editorials, abstracts, expert opinions, letters, and noncomparative studies; (2) studies reporting on robotic-assisted or 3-dimensional laparoscopic living-donor nephrectomy; (3) studies comparing laparoscopic and open nephrectomy; (4) repeated reports from authors, centers, or patient populations; or (5) considerable overlap between a cohort evaluated previously. Outcomes of interest and definitions The following outcomes were used to compare right and left laparoscopic living-donor nephrectomy: (1) donor operative parameters operative time (min), operative blood loss (ml), and warm ischemia time (min); (2) donor postoperative parameter length of hospital stay (d); (3) donor intraoperative complications bowel, liver, spleen, pleural, or lung injuries; (4) donor postoperative complications pulmonary, vascular, or urologic complications, wound infections, incisional hernias, chronic wound pain, and mortality; (5) recipient postoperative complications ureteral leak, ureteral stricture, and vascular complications; (6) conversion to open donor nephrectomy and donor blood transfusion; and (7) graft parameters delayed graft function, 1-year graft loss, and recipient creatinine levels (mg/dl) at 7 days, 1 month, and 1 year after transplant. Statistical analyses This meta-analysis followed the Quality of Reporting of Meta-Analyses guidelines and recommendations of the Cochrane Collaboration 14,15 and was performed with statistical software (Review Manager, Version 5.0, Clicktime, San Francisco, CA, USA) (Stat Manager, V4.1, Stata Corp, College Station, TX, USA). The weighted mean difference was used to analyze continuous variables, and dichotomous variables

3 216 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Exp Clin Transplant were analyzed by determination of odds ratio (OR). 16 There were 15 studies identified within 95% confidence interval (CI). The OR represented the occurrence odds of an event in the right compared with left group, and the numerical differences between the 2 groups were summarized by weighted mean difference for continuous variables. When P <.05 and 95% CI did not include the value 1, the OR and weighted mean difference were considered statistically significant. When continuous data were presented as range and mean, we used statistical algorithms and bootstrap resampling techniques to calculate and verify the standard deviations (SD). To evaluate the quality of the studies, we used the Newcastle-Ottawa Scale, with minimal modifications matched for the requirements of the present study (Table 1). 17 The quality of the studies was assessed according to the following criteria: patient selection, comparability between the 2 groups, and outcome assessment. Except for the only randomized controlled trial, each study was assessed by scoring from 0 to 9 stars. The maximum number of stars in the selection, comparability, and outcome categories was 3, 4, and 2. The studies that had 5 stars and the only randomized controlled trial were believed to be of high quality. The Higgins chi-square test was performed to evaluate heterogeneity. Statistical heterogeneity was evaluated by P and I² values. If I² < 50% and P >.1, the heterogeneity was considered to be within an appropriate range and data were pooled using a fixed-effects model. When significant heterogeneity was present (I² > 50%; P <.1), the random-effects model was applied. We performed subgroup analyses to interpret significant heterogeneity. To further investigate heterogeneity, Galbraith plots analysis was performed to identify the outliers which might have contributed to the heterogeneity. The effect of the low-quality studies on the overall effect was detected using sensitivity analysis. We performed Egger tests and Begg funnel plots to evaluate publication bias of studies in all comparison models. Results eligible studies The identified studies selected for this analysis successfully matched the selection criteria and had been published between 2000 and The search strategy generated 105 relevant clinical studies, including 19 full text articles that were further Table 1. Quality Assessment of Nonrandomized Studies Selection Comparability Outcome Assessment Total Hoda Tsoulfas Diner Omoto Gures Maartense Posselt Dol Ko Bachir Husted Ruszat Hoda Narita Checklist for quality assessment and scoring of nonrandomized studies. 1. Sample conforms to the indications for surgical treatment for renal transplant (if yes, 1 star). 2. The 2 groups are randomly assigned (if yes, 1 star; no star if the patients were selected or selection of group was not described). 3. The 2 groups are from the same sample (if drawn from the same community as the reference group, 1 star; no star if drawn from a different source or selection of group was not described). 4. Group 1: 1, age; 2, sex; 3, body mass index; 4, donor renal function; 5, number of renal vessels; 6, operating surgeon/team (if yes, 2 stars; 1 star was assigned if 1 of these 6 characteristics was not reported, even if there were no other differences between the 2 groups and other characteristics had been controlled; no star was assigned if the 2 groups differed). 5. Group 2: 7, aspirin; 8, human leukocyte antigen mismatches; 9, comorbidity; 10, related/unrelated donor/recipient; 11, race; 12, matched for follow-up (if yes, 2 stars; 1 star was assigned if 1 of these 6 characteristics was not reported, even if there were no other differences between the 2 groups and other characteristics had been controlled; no star was assigned if the 2 groups differed). 6. Monitoring indicators must be documented (yes, 1 star for information ascertained by record linkage or interview; no star if this information was not reported). 7. Follow -up not < 1 year (if yes, 1 star).

4 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: investigated. Of these, 4 studies were excluded: the data of 2 studies were incomplete, containing only averages; 1 study did not contain a control group; and the data of 1 study were unclear. Therefore, 15 studies were identified for inclusion, including 1 randomized controlled trial 18 and 14 nonrandomized comparative studies. A flow diagram outlined the process of study selection (Figure 1). Figure 2. Meta-Analysis of Right Versus Left Laparoscopic Living-Donor Nephrectomy Figure 1. Flow Diagram Outlining the Study Selection Process Study characteristics The characteristics of 15 studies that fulfilled the inclusion criteria were summarized (Table 2). Analysis was performed on 3073 patients; 2420 patients (78%) had undergone left laparoscopic living-donor nephrectomy and 653 patients (22%) had undergone right laparoscopic living-donor nephrectomy. There were 6 studies that recorded hand-assisted laparoscopic living-donor nephrectomy (total, 630 patients). There were 7 studies that recorded traditional laparoscopic living-donor nephrectomy (total, 1792 patients). There were 2 studies that recorded retroperitoneoscopic livingdonor nephrectomy (total, 651 patients). Conversion to open surgery was reported in 44 cases (1.4%) in 8 studies. There were 17 patients (0.6%) who needed donor blood transfusion in 8 studies. Data collection

5 218 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Exp Clin Transplant Table 2 Basic Characteristic of Included Studies* Study Study Country Group Sample Delayed 1-Year Donor Donor Design Size Graft Graft Loss Intraoperative Postoperative Function Complications Complications Hoda 2011 RCG Germany Left % (1/40) 2.5% (1/40) 5% (2/40) NA Right % (1/51) 1.96% (1/51) 0 NA Tsoulfas 2012 RCG Greece Left % (19/260) NA 1.2% (3/260) 6.2% (16/260) Right % (1/19) NA % (2/19) Diner 2006 RCG United States Left NA 3.1% (4/127) 3.1% (4/127) Right 40 0 NA 0 0 Minnee 2007 RCT The Left % (1/29) 3.4% (1/29) 10% (3/29) 7% (2/29) Netherlands Right % (1/31) 6.5% (2/31) 13% (4/31) 0 Omoto 2013 RCG Japan Left % (10/509) 1.8% (9/509) 0.4% (2/509) 3.1% (16/509) Right % (4/24) Gures 2013 RCG Turkey Left % (4/143) 2.8% (4/143) NA 2.1% (3/143) Right % (1/65) 1.5% (1/65) NA 3.1% (2/65) Gures 2013 RCG Turkey Left 27 NA 2.8% (4/143) NA 2.1% (3/143) Right 23 NA 8.7% (2/23) % (4/23) Maartense 2004 RCG The Left 27 NA 3.7% (1/27) 11.1% (3/27) 14.8% (4/27) Netherlands Right 23 NA 8.7% (2/23) % (4/23) Posselt 2004 RCG United States Left % (15/333) NA 0.3% (1/333) 4.2% (14/333) Right % (3/54) NA 1.9% (1/54) 1.9% (1/54) Dols 2008 RCG The Left 124 NA 0.8% (1/124) 13.7% (17/124) 9.7% (12/124) Netherlands Right 159 NA 1.3% (2/159) 5.0% (8/159) 6.9% (11/159) Ko 2008 RCG United States Left 359 NA 1.9% (7/359) 4.2% (15/359) 10.6% (38/359) Right 41 NA 4.9% (2/41) % (5/41) Bachir 2011 RCG United States Left 74 NA 2.7% (2/74) NA NA Right 20 NA 0 NA NA Husted 2005 RCG United States Left 213 NA 0 NA NA Right 40 NA 2.5% (1/40) NA NA Ruszat 2007 RCG Switzerland Left 90 NA 1.1% (1/90) 4.4% (4/90) 20% (18/90) Right 28 NA 0 7.1% (2/28) 21.4% (6/28) Hoda 2010 RCG Germany Left % (1/36) 2.8% (1/36) 5.5% (2/36) 0 Right % (1/46) 2.2% (1/46) 0 0 Narita 2006 RCG Japan Left % (8/56) NA 3.6% (2/56) 3.6% (2/56) Right % (1/12) NA 8.3% (1/12) 0 Abbreviations: NA, not available; OR, odds ratio; RCG, retrospective control group; RCT, randomized controlled trial *Dichotomous variables are presented as odds ratios. was prospective in 1 study, and the other studies were retrospective. There was 1 randomized controlled trial and the other studies were nonrandomized. Meta-analysis of right versus left laparoscopic living-donor nephrectomy For the donors, operative time was shorter in the right than left group by minutes (95% CI, to min; P =.005) (Table 3 and Figure 2). Operative blood loss was lower in the right than left group by ml (95% CI, to ml; P =.003). Comparison of warm ischemia time between the right and left group showed no significant difference (95% CI, to 0.26 min; P =.33). There was a difference in the donor intraoperative complication rate between the 2 groups (OR, 0.53; 95% CI, 0.31 to 0.92; P =.03); the left group had a higher rate of donor intraoperative complications. The intraoperative complications included bowel or liver injury, spleen or pancreas injury, pleural or lung injury, and intraoperative bleeding. However, there was no difference between groups for rate of conversion to open donor nephrectomy (OR, 0.54; 95% CI, 0.24 to 1.21; P =.14). There was no significant difference between the 2 groups in donor postoperative complication rate (OR, 1.03; 95% CI, 0.68 to 1.56; P =.88). Donor postoperative complications included pulmonary, vascular, and urologic complications, wound infection, incisional hernia, and chronic wound pain.

6 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Table 3. Meta-Analysis of Right Versus Left Group Parameters Outcome No. of No. of Odds Ratio / 95% P Heterogeneity P I² Studies Patients Weighted Mean Confidence Difference Interval Operative time (min) , % Operative blood loss (ml) , % Warm ischemia time (min) , % Donor hospital stay (d) , % Delayed graft function , % Recipient 1-year graft loss , % Recipient serum creatinine, 7 d postoperative (mg/dl) , % Recipient serum creatinine, 1 mo postoperative (mg/dl) , % Recipient serum creatinine, 1 y postoperative (mg/dl) , % Conversion to open donor nephrectomy , % Donor blood transfusion , % Donor intraoperative complications , % Donor postoperative complications , % Recipient postoperative complications , % The point estimates of the odds ratio and weighted mean difference were considered statistically significant at the level of P <.05 if the 95% confidence interval did not include the value 1. If I² < 50% and P >.1, it may be considered to indicate nonsignificant heterogeneity. Patients undergoing right or left laparoscopic livingdonor nephrectomy had similar postoperative recovery. There were no significant differences between the 2 groups regarding donor hospital stay (weighted mean difference, d; 95% CI, d to 0.05 d; P =.32). Recipient parameters Pooled analysis of 7 studies showed that there was no significant difference between the right and left group (95% CI, 0.70% to 1.41%; v =.97) in rate of overall recipient postoperative complications such as ureteral leak, ureteral stricture, or vascular complications. Analysis of 9 studies suggested that patients in the right and left groups had similar recipient delayed graft function (95% CI, 0.39% to 1.76%; P =.63). There was no significant difference in the rate of recipient 1-year graft loss (95% CI, 0.71% to 3.00%; P =.3) from the analysis of 11 studies. Both groups had similar graft function assessed at 7 days, 1 month, and 1 year using serum creatinine levels (mg/dl) (7 d: 95% CI, to 0.09; P =.54) (1 mo: 95% CI, to 0.06, P =.35) (1 y: 95% CI, to 0.16, P =.73). Overall, meta-analysis of related data showed no significant differences between the right and left groups in recipient relative parameters. Subgroup analysis of right versus left handassisted surgery There was no significant change in the results for most outcomes from the original analysis (Figure 3). The operative blood loss was similar (right, 21.4 ml; left, 6.7 ml; 95% CI, ml to ml; P =.41) and the rate of donor intraoperative complications was similar (right, 27.9%; left, 18.5%; 95% CI, 0.23% to 1.46%; P =.25) between the 2 groups. Recipient serum creatinine at 7 days after surgery was not compared between the 2 groups because it was included only in 1 study. Right versus left living-donor nephrectomy This comparison revealed similar results as in the original analysis (Figure 4). There was a difference in the rate of donor conversion to open living-donor nephrectomy between the 2 groups (right, 41.9%; left, 36.8%; 95% CI, 0.11% to 0.97%; P =.04). There was no difference in the rate of donor blood transfusion between right and left transperitoneal laparoscopic living-donor nephrectomy (95% CI, 0.06% to 1.79%; P =.20). Right versus left retroperitoneoscopic nephrectomy There was no significant change in the results for most outcomes from the original analysis (Figure 5). However, compared with the original analysis, there were no differences in operative time (95% CI, min to min; P =.77) and operative blood loss (95% CI, ml to ml; P =.47). Sensitivity analysis There were a total of 7 high quality studies that achieved 5 stars according to the modified Newcastle-Ottawa scale and were included in the

7 220 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Exp Clin Transplant Figure 3. Subgroup Analysis of Right Versus Left Hand-Assisted Living-Donor Nephrectomy sensitivity analysis; the only randomized controlled trial also was included in the sensitivity analysis. The results of sensitivity analyses from the above studies revealed no significant differences from the overall analysis (Table 4, Figure 6). Heterogeneity analysis For the right and left laparoscopic living-donor nephrectomy, the I² value of heterogeneity was > 50% and P <.10 for operative time (I² = 86%; P <.00001), warm ischemia time (I² = 68%; P =.0004), and postoperative recipient serum creatinine at 1 year (I² = 79%; P =.0009), which indicated statistically significant heterogeneity between studies. To explore the sources of heterogeneity, we performed subgroup analyses and sensitivity analysis. We conducted subgroup analyses according to the specific operative method. In the hand-assisted laparoscopic livingdonor nephrectomy, heterogeneity of operative time (I² = 56%; P =.05), warm ischemia time (I² = 56%; P =.06), and postoperative recipient serum creatinine at 1 year (I² = 0%; P =.66) decreased significantly, but in the other 2 subgroups, heterogeneity was not reduced. To further investigate the heterogeneity, we performed Galbraith plot analysis to identify outliers that might have contributed to the heterogeneity. The studies by Posselt and coworkers and Dols Table 4. Sensitivity Analysis Outcome No. of No. of Odds Ratio / 95% P Heterogeneity P I² Studies Patients Weighted Mean Confidence Difference Interval Operative time (min) , % Operative blood loss (ml) , % Warm ischemia time (min) , % Donor hospital stay (d) , % Delayed graft function , % Recipient 1-year graft loss , % Recipient serum creatinine, 7 d postoperative (mg/dl) , % Recipient serum creatinine, 1 mo postoperative (mg/dl) , % Recipient serum creatinine, 1 y postoperative (mg/dl) , % Conversion to open donor nephrectomy , % Donor blood transfusion , % Donor intraoperative complications , % Donor postoperative complications , % The point estimates of the odds ratio and weighted mean difference were considered statistically significant at the level of P <.05 if the 95% confidence interval did not include the value 1. If I² < 50% and P >.1, it may be considered to indicate nonsignificant heterogeneity.

8 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Figure 4. Subgroup Analysis of Right Versus Left Living-Donor Nephrectomy and associates were outliers in operative time; I² decreased and P >.10 (I² = 46%; P =.05) after excluding these 2 studies from the operative time model. Galbraith plot analysis showed that the studies by Dols and associates and Maartense and coworkers were outliers in warm ischemia time; I² decreased to < original value and P > original value after excluding these 2 studies (I² = 54%; P =.02). In the model for postoperative recipient serum creatinine at 1 year, the number of studies was insufficient and we could not identify the outliers that might have contributed to heterogeneity. Publication bias The funnel plots were calculated based on operative time, warm ischemia time, and postoperative recipient serum creatinine at 1 year (Figure 7). All trials were within 95% CI limits, and no evidence of publication bias was noted. The Egger test was performed to present statistical evidence about the symmetry of funnel plots. There was no evidence of publication bias in terms of operative time (P =.84), warm ischemia time (P =.406), or postoperative recipient serum creatinine at 1 year (P =.843).

9 222 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Exp Clin Transplant Figure 5. Subgroup Analysis of Right Versus Left Retroperitoneoscopic Living-Donor Nephrectomy Figure 6. Sensitivity Analysis

10 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Figure 6. Sensitivity Analysis (Continued) Figure 7. Funnel Plots Based on Outcomes A. The Funnel Plot of Recipient Serum Creatinine (1 yr Postop) B. The Funnel Plot of the Operation Time C. The Funnel Plot of Warm İschemia Time Abbreviations: MD, mean difference; SE, effect estimate versus effect estimate for each study under the outcome Studies are marked by a dot and 95% confidence intervals by lines. Discussion Many studies suggested that laparoscopic nephrectomy in living-donor transplant was a safe alternative to the open technique. 3,19 More medical centers performed the laparoscopic technique. But urologists could not decide which side was better, only from their experience and intuition. This systematic analysis was essential. Our study was a relevant meta-analysis of all the available literature comparing the left and right sides in laparoscopic living-donor nephrectomy. In the present study, the right group had shorter operative time of median minutes and lower

11 224 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Exp Clin Transplant operative blood loss of ml compared with the left group. In 1 study, the ranges for operative blood loss were significantly wider (left, 10 to 1000 ml; right, 25 to 1200 ml); this might relate to the unpredictability and uncertainty of the operative process. The ml difference in donor blood loss and min difference in operative time were statistically significant but most likely were not clinically relevant. Therefore, we do not need to further study the origin of the subtle differences in the operation. In our study, there were no obvious differences regarding donor or recipient overall postoperative complications. However, the outcomes showed that the left group had a higher rate of donor total intraoperative complications than the right group (OR, 0.53; P =.03). The donor intraoperative complications included bowel or liver injury, spleen or pancreas injury, pleural or lung injury, and vascular injury. The more complex anatomic structures surrounding the left kidney might contribute to the higher rate of overall donor intraoperative complications. Although conversion to open technique was not desired, it sometimes was inevitable. The intraoperative complications were the chief causes of conversion to open donor nephrectomy. 20 Nevertheless, our study showed no significant differences in the rate of conversion to open donor nephrectomy or donor blood transfusion between the 2 groups; the higher incidence of intraoperative complications on the left side did not lead to a higher incidence of open conversion. Thus, intraoperative complications on the left side were not severe, and urologists may provide nonoperative treatment. We considered that the more complex anatomic structures surrounding the left kidney may cause some but not serious intraoperative effects on the operation because the longer renal vein of the left side may make the implant procedure less demanding and may reduce the difficulty of the operation Therefore, in our analysis, the right side did not have an advantage regarding the rate of overall donor intraoperative complications compared with the left side. For urologists, early recovery of graft function is most concerning. 24 Our study showed no differences in the warm ischemia time. A related survey by Abreu and coworkers 25 reported that warm ischemia time plays an important role in delayed graft function. In our analysis, the rate of delayed graft function was similar in the right and left groups (right, 3.2%; left, 4.1%). Studies comparing left versus right laparoscopic nephrectomy demonstrated no difference in terms of donor hospital stay. In the analysis of follow-up data, no significant differences in postoperative measures of graft function between the left and right groups were evident such as postoperative recipient serum creatinine at 7 days, 1 month, or 1 year. Some studies showed that serum creatinine was useful in detecting minor degrees of deterioration of renal function, and glomerular filtration rate may be estimated accurately based on serum creatinine according to the Kidney Disease Improving Global Outcomes and National Kidney Foundation/Kidney Disease Outcomes Quality Initiative guidelines. 26,27 Nevertheless, as a more intuitive indicator, the rate of 1-year graft loss should be evaluated. In our meta-analysis, the rates of 1-year graft loss on the right and left sides were 2.6% and 2%, and we found no statistical significance in the comparison of the rate of 1-year graft loss. The sensitivity analyses were conducted after excluding the low-quality trials from the nonrandomized controlled trials. We compared the effect on the results of the overall analyses. The results of sensitivity analyses were consistent with the results of the overall analysis; no significant differences were revealed in all indicators from the overall results. This increases the credibility of our overall analysis. Heterogeneity analysis of the right and left laparoscopic living-donor nephrectomy suggested significant heterogeneity in operative time, warm ischemia time, and postoperative recipient serum creatinine at 1 year. To explore the sources of heterogeneity, we performed subgroup analyses according to the specific operative method. In handassisted nephrectomy, the heterogeneity of 3 models decreased significantly, but in the other 2 subgroups, heterogeneity did not decrease. Therefore, the specific operative method was not the main source of heterogeneity. To further investigate the heterogeneity, we performed Galbraith plot analysis to identify the outliers that might have contributed to the heterogeneity. Our results showed that the studies of Posselt and associates and Dols and coworkers were outliers in operative time. All I² values decreased and P was >.10 after excluding these studies in the operative time model. Galbraith plot analysis showed that the studies by Dols and associates and Maartense and coworkers were outliers in warm ischemia time. The I² value

12 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: decreased < original value and P was > original value after excluding the studies of Dols and coworkers and Maartense and associates. In the model for postoperative recipient serum creatinine at 1 year, because the number of studies included was not sufficient, we could not identify the outliers that might have contributed to the heterogeneity. The results indicated that these studies might be the major source of heterogeneity for operative time and warm ischemia time. Many studies were affected by factors such as patient preference and could not randomly assign the operative approach. Therefore, nonrandomized controlled trials accounted for a high proportion in our analysis, which is a possible criticism that the meta-analysis may reinforce inherent systematic biases of the studies, cause spurious statistical stability, and discourage further research. However, Deeks and coworkers 28 have assessed nonrandomized controlled trials by using resampling techniques; they showed that results of randomized and nonrandomized studies may differ, but similarities and differences may be explained by other confounding factors. Therefore, we believe that meta-analysis of nonrandomized studies may be useful when randomized controlled trials are not available. Nevertheless, further prospective randomized trials are required for a more comprehensive comparison of the 2 sides. In several studies, ranges but not standard deviations were recorded. Some workers doubt that this may lead to publication bias. However, a recent study by Hozo and coworkers 29 provided a rigorous scientific method to estimate the standard deviation. Using these formulas, they suggested that we can use clinical trials even when not all of the information is available and/or reported. In addition, in our meta-analysis, studies that recorded ranges rather than the standard deviation only accounted for a small proportion. The results of Egger test did not suggest any evidence of publication bias in the overall outcomes. In some studies, there was slight variability in terms of definitions, inclusion criteria, and measurement of outcomes. This variability might introduce bias into the current study. However, consensus in definitions, inclusion criteria, and measurement of outcomes was reached between reviewers by consulting relevant experts and counterparts. The analysis results showed little heterogeneity for most of the continuous outcomes and all dichotomous outcomes examined in this analysis. Limitations of the present study included the differences in the number of patients between the 2 groups. Many urologists prefer the left than right side due to the bias of the less demanding operation from the longer renal vein, and they conclude by intuition that the less demanding operation may improve the effect of the operation. Therefore, the number of cases was less in the right than left group. More trials of the right procedure are required for a more comprehensive comparison of the 2 sides. In our study, right laparoscopic living-donor nephrectomy might be associated with shorter operative time, lower operative blood loss, and fewer donor intraoperative complications. We considered that several modifications used for right donor nephrectomy and recipients played an important role in improving the safety and efficacy of right laparoscopic living-donor nephrectomy by prolonging the vein length with inverted kidney transplant, 30 the use of hand-assisted devices, and use of a modified Satinsky atraumatic vascular clamp. 31 However, compared with left laparoscopic living-donor nephrectomy, the modifications did not decrease the operative difficulty significantly and made the operation less demanding for the urologists. In summary, right laparoscopic living-donor nephrectomy did not have significant clinical advantages. Moreover, there were no differences between left and right laparoscopic living-donor nephrectomy in most important outcomes. Our meta-analysis statistically confirmed that right and left laparoscopic living-donor nephrectomy were similar in the effects of surgery and postoperative graft function. Therefore, when there are no differences in bilateral renal function, surgeons may take the right or left kidney randomly from the perspective of intraoperative and postoperative effects. However, we recommend using the left kidney because the longer renal vein of the left kidney may reduce operative difficulty in taking the donor kidney and may make the operation less demanding for urologists. References 1. Pierce GA, Graham WK, Kauffman HM Jr, Wolf JS. The United Network for Organ Sharing: 1984 to Transplant Proc. 1996;28(1):12-15.

13 226 Kun Wang et al/experimental and Clinical Transplantation (2015) 3: Exp Clin Transplant 2. Harper AM, Taranto SE, Edwards EB. The OPTN waiting list, Clin Transpl. 2002: Ratner LE, Ciseck LJ, Moore RG, Cigarroa FG, Kaufman HS, Kavoussi LR. Laparoscopic live donor nephrectomy. Transplantation. 1995;60(9): Tooher RL, Rao MM, Scott DF, et al. A systematic review of laparoscopic live-donor nephrectomy. Transplantation. 2004;78(3): Matas AJ, Bartlett ST, Leichtman AB, Delmonico FL. Morbidity and mortality after living kidney donation, : survey of United States transplant centers. Am J Transplant. 2003;3(7): Kok NF, Weimar W, Alwayn IP, Ijzermans JN. The current practice of live donor nephrectomy in Europe. Transplantation. 2006;82(7): Lind MY, Hazebroek EJ, Hop WC, Weimar W, Jaap Bonjer H, IJzermans JN. Right-sided laparoscopic live-donor nephrectomy: is reluctance still justified? Transplantation. 2002;74(7): Merlin TL, Scott DF, Rao MM, et al. The safety and efficacy of laparoscopic live donor nephrectomy: a systematic review. Transplantation. 2000;70(12): Mandal AK, Cohen C, Montgomery RA, Kavoussi LR, Ratner LE. Should the indications for laparascopic live donor nephrectomy of the right kidney be the same as for the open procedure? Anomalous left renal vasculature is not a contraindiction to laparoscopic left donor nephrectomy. Transplantation. 2001;71(5): Ratner LE, Kavoussi LR, Chavin KD, Montgomery R. Laparoscopic live donor nephrectomy: technical considerations and allograft vascular length. Transplantation. 1998;65(12): Buell JF, Edye M, Johnson M, et al. Are concerns over right laparoscopic donor nephrectomy unwarranted? Ann Surg. 2001;233 (5): Gill IS, Uzzo RG, Hobart MG, Streem SB, Goldfarb DA, Noble MJ. Laparoscopic retroperitoneal live donor right nephrectomy for purposes of allotransplantation and autotransplantation. J Urol. 2000;164(5): Narita S, Inoue T, Matsuura S, et al. Outcome of right hand-assisted retroperitoneoscopic living donor nephrectomy. Urology. 2006;67 (3): Clarke M, Horton R. Bringing it all together: Lancet-Cochrane collaborate on systematic reviews. Lancet. 2001;357(9270): Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283(15): DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3): Athanasiou T, Al-Ruzzeh S, Kumar P, et al. Off-pump myocardial revascularization is associated with less incidence of stroke in elderly patients. Ann Thorac Surg. 2004;77(2): Minnee RC, Bemelman WA, Maartense S, Bemelman FJ, Gouma DJ, Idu MM. Left or right kidney in hand-assisted donor nephrectomy? A randomized controlled trial. Transplantation. 2008; 85(2): Kokkinos C, Nanidis T, Antcliffe D, Darzi AW, Tekkis P, Papalois V. Comparison of laparoscopic versus hand-assisted live donor nephrectomy. Transplantation. 2007;83(1): Eng M. The role of laparoscopic donor nephrectomy in renal transplantation. Am Surg. 2010;76(4): Leventhal JR, Deeik RK, Joehl RJ, et al. Laparoscopic live donor nephrectomy - is it safe? Transplantation. 2000;70(4): Wolf JS Jr, Marcovich R, Merion RM, Konnak JW. Prospective, case matched comparison of hand assisted laparoscopic and open surgical live donor nephrectomy. J Urol. 2000;163(6): Hsu JW, Reese PP, Naji A, Levine MH, Abt PL. Increased early graft failure in right-sided living donor nephrectomy. Transplantation. 2011;91(1): Jayaram D, Kommareddi M, Sung RS, Luan FL. Delayed graft function requiring more than one-time dialysis treatment is associated with inferior clinical outcomes. Clin Transplant. 2012;26 (5):E536-E Abreu SC, Goldfarb DA, Derweesh I, et al. Factors related to delayed graft function after laparoscopic live donor nephrectomy. J Urol. 2004;171(1): Spanaus KS, Kollerits B, Ritz E, et al. Serum creatinine, cystatin C, and beta-trace protein in diagnostic staging and predicting progression of primary nondiabetic chronic kidney disease. Clin Chem. 2010;56(5): Tidman M, Sjöström P, Jones I. A comparison of GFR estimating formulae based upon s-cystatin C and s-creatinine and a combination of the two. Nephrol Dial Transplant. 2008;23(1): Deeks JJ, Dinnes J, D'Amico R, et al. Evaluating non-randomised intervention studies. Health Technol Assess. 2003;7(27):iii-x, Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol. 2005;5: Simforoosh N, Aminsharifi A, Tabibi A, Fattahi M, Mahmoodi H, Tavakoli M. Right laparoscopic donor nephrectomy and the use of inverted kidney transplantation: an alternative technique. BJU Int. 2007;100(6): Simforoosh N, Soltani MH, Basiri A, et al. Evolution of laparoscopic live donor nephrectomy: a single-center experience with 1510 cases over 14 years. J Endourol. 2014;28(1):34-39.