Not-so-simple nephrectomy: Comparative analysis of radical and simple nephrectomy in a high-volume tertiary referral center.

N/A.

Ariadni Papadopoulou: Investigation; Writing—original draft; Methodology; Validation; Visualization; Writing—review & editing; Software; Data curation; Project administration; Formal analysis. Nicholas Campain: Investigation; Writing—original draft; Methodology; Validation; Writing—review & editing. Yasmin Abu‐Ghanem: Writing—review & editing; Methodology; Investigation; Validation. Nimlan Shanmugathas: Data curation; Writing—review & editing; Methodology; Validation. Marios Poullas: Formal analysis; Software; Methodology; Validation; Investigation. Faiz Mumtaz: Investigation; Writing—review & editing; Validation; Methodology. Ravi Barod: Investigation; Validation; Methodology; Writing—review & editing; Supervision. Maxine Tran: Investigation; Validation; Writing—review & editing; Methodology. Axel Bex: Investigation; Methodology; Validation; Writing—review & editing. Prasad Patki: Conceptualization; Investigation; Writing—original draft; Methodology; Validation; Writing—review & editing; Supervision.

After obtaining institutional review board approval, we retrospectively analyzed a series of patients undergoing minimally invasive and open nephrectomy in a single institution. Three hundred and forty‐four RNs (January 2018–October 2020) and 150 SNs (January 2016–October 2020) performed in a specialist kidney cancer center, were included in the study. Data on SN were collected over a longer timeframe due to preponderance of RN in a tertiary renal cancer unit to perform a balanced statistical analysis. Partial nephrectomies were excluded. Baseline data recorded to assess population comparability included age, gender, laterality, BMI, INR, intraoperative creatinine, eGFR, American Society of Anesthesiologists (ASA) score, and indication for surgery. Data were retrospectively retrieved through the prospective patient database, operative notes, inpatient records, discharge summaries, and clinic correspondence. Radiographic tumor size was retrieved from radiology reports. For subgroup analysis, surgical techniques were grouped into minimally invasive (MI‐robotic and laparoscopic) and open surgery and outcomes for SN and RN were compared by technique. RN were classified according to tumor size into three groups: Group 1 cT1, Group 2 cT2a, and Group 3 cT2b‐T3 as per the TNM classification 7 and individually compared to SN. SN was classified further according to pathology as Infected, including recurrent UTIs, abscesses, pyelonephritis, and XPN and non‐infected including atrophic or non‐functioning kidneys secondary to PUJO/hydronephrosis, retroperitoneal fibrosis, pelvic and retro‐vaginal endometriosis, calculus without documented infection, polycystic kidney disease, and iatrogenic trauma and compared. Outcome variables included operative time, length of hospital stay, estimated blood loss, conversion to open surgery, planned and planned intensive care unit (ICU)/high dependency unit (HDU) stay, readmission rates within 28 days postoperatively, intraoperative, and postoperative complications. Operative time was defined as the total interval between skin incision and skin closure. Postoperative complication up to 28 days postoperatively and their management were recorded and classified according to the Clavien–Dindo classification as minor (Clavien grade 1–2) or major (Clavien grade 3–5) complications. 8 , 9 Descriptive statistics of demographics and clinical characteristics including median, and interquartile range were calculated accounting for the non‐normal data distribution. Univariate analysis utilized the Mann–Whitney U test for quantitative variables, Chi‐squared/Fisher's exact test for nominal categorical variables, and Mantel–Haenszel test of trend for overall postoperative complications. Notice that the Mann–Whitney U test used due to non‐normal data distribution does not compare medians and instead considers all datapoints. It tests for the null hypothesis that when two values X  and Y , one from each population, are selected and compared there is equal probability that X is greater than Y as that Y is greater than X . Hence, statistically significant difference can be observed with similar medians. In those cases, means are reported for completion. Multiple linear regression evaluated the independent effect of operation type (SN vs. RN) on perioperative outcomes across all cases after adjusting for age, gender, BMI, preoperative eGFR, laterality, ASA score, and surgical approach. Adjusted beta coefficients, standard error, test statistic, and p ‐values were recorded. For all tests, a two‐sided p ‐value <0.05 was considered statistically significant. Nearest neighbor analysis was performed and missing data were completed based on the 23 closest datapoints available. The analysis was carried out with SPSS 10 and Matlab. 11 These data were presented at the BAUS 2021 “Global Urology” Virtual Meeting (P12‐9) and the European Association of Urology 2021 Virtual Conference (abstract number: AM21‐4735).

A total of 344 RNs and 129 SNs were included. Patient demographics are summarized in Table  1 . Patient and clinical characteristics in RN and SN groups ( N  = 473). Note : p ‐value < 0.05 was considered significant. p ‐values presented in bold are statistically significant. p  < 0.05. p  < 0.001. SN was associated with significantly younger age at time of surgery (63 vs. 48.5; p  < 0.001), female gender (58% vs. 36%; p  < 0.001) and lower ASA score ( p  = 0.038) and eGFR ( p  = 0.032). MIS approach was used for 91% of RN and 95% of SN ( p  = 0.127). Robotic cases accounted for 60% of RN and 53% of SN. In the univariate analysis, SN was associated with significantly higher operative time compared to RN (mean: 136.42 min vs. 126.91 min, p  = 0.023; Table  2 ; Figure  1a ). The difference remained significant in multivariate analysis ( p  = 0.001; Table  3 ). Factors associated with longer operative time on multivariate analysis included higher BMI ( p  < 0.001), ASA score 3 (vs. ASA 1) ( p  = 0.036), and open surgical approach ( p  < 0.001; Table  3 ). Perioperative outcomes for all cases in the RN and SN groups ( N  = 473). Minor complications (Clavien grade 1–2) Major complications (Clavien grade 3–5) Note : Significance may be observed where median values are similar due to differences in data distribution between the groups. The Mann–Whitney U test compares rank sums. p ‐value <0.05 was considered significant. p ‐values presented in bold are statistically significant. p  < 0.05. p  < 0.01. p  < 0.001. Data distribution for select intra‐ and postoperative outcomes of radical (RN) and simple nephrectomies (SN). (a) Box plot of operative time across all cases. (b) Box plot of length of stay across all cases. The median length of stay is equal to the 25th percentile for both RN and SN (2 nights). (c) Box plot of operative time across all cases classified by RN tumor size. (d) Bar graph of postoperative complication frequency in SN and RN according to tumor size. Multiple linear regression of select outcomes after SN and RN ( n  = 473). Note : p ‐value < 0.05 was considered significant. p ‐values presented in bold are statistically significant. p  < 0.05. p  < 0.01. p  < 0.001. Blood loss did not differ significantly between the groups ( p  = 0.482) and the only independent predictors of increased blood loss were higher BMI ( p  = 0.026) and open surgical approach ( p  < 0.001; Table  3 ). There were 9 (3%) intraoperative complications in the RN and 5 (4%) in the SN group ( p  = 0.371). Conversions to open were noted in eight RN (2%) and three SN (2%) with no significant difference between the groups ( p  = 1.000; Table  2 ). There were 23 (7%) HDU/ICU admissions in the RN and seven (6%) in the SN group with no significant difference ( p  = 0.729; Table  2 ). Ten (43%) RN HDU/ICU admissions were planned compared to only two SN HDU/ICU admissions (29%) ( p  = 0.481). Independent predictors associated with increased risk of HDU/ICU admission were open surgical approach ( p  < 0.001) and ASA score 3 ( p  = 0.007; Table  3 ). Despite readmission rates being higher in the SN cohort (6% vs. 2%) due to higher rates of wound infection, acute kidney injury, and postoperative ileus, the difference was not statistically significant ( p  = 0.064; Table  2 ). The postoperative complication rate following SN was significantly higher than in the RN group in univariate analysis (39% vs. 21%; p  = 0.006, Table  2 ). SN remained an independent predictor of increased postoperative complications on multivariate analysis ( p  < 0.001, Table  3 ). ASA score 3 ( p  = 0.002) and open surgical approach ( p  < 0.001) were also independent predictors of increased postoperative complications (Table  3 ). Minor postoperative complication rate was significantly higher in the SN group in univariate (35% vs. 17%; p  < 0.001, Table  2 ) and multivariate analysis ( p  = 0.001, Table  3 ) with ASA score 3 being the other independent predictor of minor complications ( p  = 0.030). Postoperative neuropraxia‐related complications such as numbness and referred pain were significantly more common in the SN cohort (10 vs. 2 events; p  < 0.001). Major postoperative complication rates remained low in both groups (3% vs. 4%, p  = 0.374) and were associated with lower eGFR ( p  = 0.049) and open surgical approach ( p  = 0.020) on multivariate assessment (Table  3 ). Only one perioperative mortality was noted in the SN group. The patient suffered from ischemic colitis, subsequently succumbing to complications. Detailed postoperative complications are displayed in Table  S1 . The mean length of stay was longer in the SN cohort (mean: 3.43 vs. 2.92 days) (Table  2 ; Figure  1b ). While this difference was not significant in univariate analysis, operation type was recognized as a significant predictor of length of stay in the multivariate model ( p  = 0.049). Low eGFR ( p  = 0.011), ASA score 3 ( p  = 0.002), and open surgical approach ( p  < 0.001) were independent predictors of prolonged length of stay (Table  3 ). When comparing the SN and RN Group 1 (cT1) the SN cases were associated with significantly higher operative time (Mean: 136 min vs. 117 min; p  < 0.001), length of stay (Median = 3.43 vs. 2.69 days; p  = 0.014) and more overall postoperative complications (39% vs. 17%; p  < 0.001; Figure  1c ). Clavien 1–2 complications were significantly more common in the SN group than in RN Group 1 (35% vs. 16%; p  < 0.001) but no difference was noted in major complications. When comparing the SN and RN Group 2 (cT2a), SN was associated with significantly higher rates of Clavien 1–2 complications (35% vs. 16%; p  = 0.001; Figure  1d ). RN Group 3 (cT2b‐T3) had significantly higher blood loss than SN (mean: 343 mL vs. 75 mL; p  < 0.001; Table  4 ; Table  S2 ). Perioperative outcomes of SN and RN groups according to tumor size. Note : Significance may be observed where median values are similar due to differences in data distribution between the groups. The Mann–Whitney U test compares rank sums. p ‐value <0.05 was considered significant. p ‐values presented in bold are statistically significant. p  < 0.05. p  < 0.01. p  < 0.001. For MIS approach, SN cases were associated with significantly higher operative time ( p  = 0.010) and length of stay ( p  = 0.028) compared to MI RN (Table  2 ). Minor complications following MIS SN were significantly higher than in the MIS RN group (31% vs. 15%; p  < 0.001), however, major complication rates were similar (Table  5 ; Table  S3 ). Subgroup analysis of select perioperative outcomes by surgical approach and SN pathology. Note : p ‐value < 0.05 was considered significant. p ‐values presented in bold are statistically significant. p  < 0.05. p  < 0.01. p  < 0.001. Open RN were associated with significantly higher median blood loss than open SN cases (Table  5 ; Table  S3 ). In the open SN group, higher proportion of minor complications (50% vs. 26%) were observed while major complications were more frequent in the open RN cohort (6% vs. 0%) but the trends did not reach significance ( p  = 0.341; p  = 1.000). Of the 129 SNs, 50 (38.8%) were classified as infected and 79 (61.2%) non‐infected (Table  S4 ). Infected SN was associated with significantly higher operative time (150 vs. 120; p  < 0.001), length of stay (3 vs. 2; p  = 0.005), blood loss (50 vs. 20; p  = 0.016), and unplanned HDU/ICU admissions (6 vs. 1; p  = 0.019; Table  5 ; Table  S5 ). There was no significant difference in intraoperative and overall postoperative complications, open conversions, and readmissions between the groups.

This study was approved by the Local Reviewer Board of Royal Free Hospital, London, United Kingdom.

N/A.

N/A.

In this large single‐center study, we have compared perioperative outcomes of SNs and RNs. Our cohort contains the largest collection of robotically assisted operations and the first subgroup analysis of RN on T2 and T3 RCC in the literature (Table  6 ). A summary of negative intra‐ and postoperative outcome predictors with corresponding recommendations to mitigate them is displayed in Table  7 . Summary of literature comparing SN and RN intraoperative and postoperative outcomes. Abbreviations: RN, radical nephrectomy; SN, simple nephrectomy. No available information on radical nephrectomy tumor stage. Independent negative predictors of intraoperative and postoperative outcomes in SN and RN with corresponding recommendations to mitigate negative outcomes based on the authors' experience. Open surgical technique (vs. MI) ASA score 3 (vs. 1) Higher BMI Infected pathology Open surgical technique (vs. MI) ASA score 3 (vs. 1) Decreased eGFR Infected pathology Open surgical technique (vs. MI) Higher BMI Infected pathology Open surgical technique (vs. MI) Decreased eGFR Abbreviations: ASA, American Society of Anesthesiologists; BMI, body mass index; eGFR, estimated Glomerular Filtration Rate; HDU, high dependency unit; ICU, intensive care unit; MI, minimally invasive; RN, radical nephrectomy; SN, simple nephrectomy. Patients in the SN group were younger, more often female, with a lower ASA score. This may reflect higher rates of recurrent UTI and pyelonephritis in younger women and early presentation with non‐functioning kidney due to long‐standing PUJO in younger population. Atrophic kidneys in SN cohort accounts for lower pre‐operative eGFR. Renal cancer is commonly associated with obesity, 12 hence lower BMI in the SN group is not surprising. 13 , 14 SN cases were associated with significantly higher operative time, as previously reported. 15 , 16 This may be attributed to inflammatory adhesions, fibrotic and dense tissue, and infection with edematous friable tissue making intraoperative dissection and anatomical structure differentiation challenging, leading to longer operative duration. However, the median operative time was similar in SN for infected pathology and RN in T2b and above disease. MI approach was an independent predictor of decreased operative time as previously reported, 17 likely due to case selection bias. When comparing MI SN and RN, SN cases were still associated with significantly higher operative time. Zelhof et al. report a higher conversion rate to open surgery in laparoscopic SN compared to RN. 4 Permpongkosol et al. have also reported higher conversion rates for SN. 18 In our patient cohort, there was no significant difference in conversion to open surgery between the MI SN and MI RN groups (2% vs. 3%), with the majority of MI SN performed robotically compared to the BAUS nationwide dataset (53% vs. 0%, respectively). However, our overall postoperative SN complication rate across all Clavien–Dindo grades was 35% and higher than that reported by BAUS and Keeley et al. (11.9% and 18%, respectively). 4 , 19 This might be due to centralization of technically difficult cases and high‐morbidity patients declined surgery elsewhere. Estimated blood loss did not significantly differ between RN and SN. Open surgical approach was an independent predictor of increased blood loss, with increased blood transfusion rates reported in literature. 20 Higher BMI was also associated with increased blood loss as previously reported 15 due to increased visceral fat resulting in significantly longer operative time. 15 As such, area of visceral fat measured by CT scan provides a more accurate blood loss measure than BMI. 21 RN for cT2b and above had significantly higher blood loss than SN. Tumor size >10 cm and venous involvement was the critical point beyond which RNs become technically more challenging than SNs. The only independent ICU admission predictors were open approach and ASA score 3, both associated with increased risk of postoperative complications and ICU admissions in RN in the literature. 22 , 23 However, only 2 of 7 ITU admissions in SN group were planned indicating an unanticipated need for higher monitoring and support in SN postoperatively. There was a significantly higher operative time, length of stay, blood loss, and HDU/ICU stay observed in our infected SN cohort compared to non‐infected SN group. SN (vs. RN) was an independent predictor of higher postoperative complications even after adjusting for differences in age, gender, BMI eGFR, ASA score, and surgical approach between the groups. Keshavamurthy et al. also reported a higher postoperative complication rate of 32.9% in SN patients compared to 25% in RN with a preponderance of minor complications, although the results did not reach significance. 16 ASA score 3 was an independent predictor of increased minor complications on a background of patient comorbidities as discussed in other publications. 24 , 25 More patients in the SN cohort experienced postoperative neuropathy‐related complications such as numbness and referred pain than in the RN cohort ( p  < 0.001). Although reversible, this finding should be communicated to patients to better inform counseling during the consent process. A possible explanation is genitofemoral nerve injury. 26 , 27 , 28 Difficult posterior renal plane dissection and trauma to the genitofemoral nerve may be more likely during SN due to adherent or inflammatory tissue around the psoas muscle perimysium, for example, in patients with XPN or emphysematous pyelonephritis with multiple percutaneous drainage tube insertions. Open surgical approach was an independent predictor of postoperative complications, including major complications. Zelhof et al. reported higher complication rates in open SN, 4 however, no analysis of significance was performed making open surgery a novel independent predictor of major postoperative complications in our study. The selection of challenging cases for open nephrectomy may justify this. However, frequent utilization of robotic interface and low conversion rates limit the use of open technique leading to reduced opportunities for surgical training. To our knowledge, this is first report comparing SN outcomes with RN for T1, T2a, and T2b‐T3 RCC. SN was associated with significantly higher operative time, length of stay, and Clavien 1–2 complications compared to RN for T1 tumors as reported by Zelhof et al. 4 When compared with RN for cT2a, the SN group still had significantly higher Clavien 1–2 complication rate. Our study is limited by its retrospective design. However thorough recording of complications was achieved by using a prospectively collected database and confirmed by sourcing multiple documents pertaining to the patient pathway. Any complications in patients repatriated to referring trusts were actively recorded and those needing further treatments were transferred to the center. In conclusion, despite SN being carried out in younger patients, they have longer operative times, LOS and Clavian 1–2 complications compared to RN for cT1 tumors. Although the postoperative complication rate is higher in SN compared to cT2a RN, RN for tumors greater than 10 cm or for cT2b and above stage have higher blood loss. The overall readmission rates and unplanned transfers to HDU remain high in SN compared to all RN. MI surgery (predominantly robotic) remains a safe option for SN in infected or non‐infected pathology with minimal conversion to open. When compared to RN, SN carries similar or greater surgical morbidity to patients, hence postoperative intensive care and enhanced recovery planning may be essential, particularly with infected pathology.

Simple nephrectomy (SN) is performed for the treatment of non‐functioning kidneys, symptomatic and recurrent infection, and associated sequalae such as abscess formation and fistulization. 1 , 2 Evidence suggests “simple” nephrectomy is far from simple. 1 , 3 , 4 Inflammatory and fibrotic tissue reaction, such as those associated with xanthogranulomatous pyelonephritis (XPN), can lead to difficulty with surgical tissue planes, making surgery challenging. A contemporary analysis of the British Association of Urological Surgeons (BAUS) national dataset found SN cases to have 1.8 times higher risk of conversion‐to‐open ( p  = 0.0039) and significantly increased rate of transfusions (4.8 vs. 2.8%; p  = 0.0143) compared to T1 radical nephrectomies (RNs). Intra‐ and postoperative complications were higher in SN cohort (5.2 vs. 3.7% and 11.9 vs. 10%), although the relationship did not reach significance. 4 Conversely, an online survey of ( n  = 95) urologists reported no difference between urologists' views of outcomes and complications of SN and RN. 5 Furthermore, Sahai et al. found no significant difference in mental health‐related quality of life of RN and SN patients. 6 However, there are no reports comparing SN outcomes with outcomes of RN on T2 and T3 renal cell carcinoma (RCC) and no relevant reports including robotic cases or adjusting for confounders with multivariate analysis. To elucidate whether SN is a negative predictor of perioperative outcomes and complications, we retrospectively compared the intra/peri‐operative outcomes of all SN and RN in a high‐volume institution. This report is the first to adjust for discrepancies in patient demographics and surgical approach between SN and RN with a multivariate investigation. It also includes the first subgroup analysis comparing RN on T2a and T2b‐T3 disease and the only collection of robotically assisted cases in relevant literature, facilitating comparison of SN and RN outcomes within evolving surgical practices.

The authors declare no conflict of interest.

Table S1–S5