Evaluation of predictive factors for i-CLARAS (intraoperative complications in laparoscopic renal and adrenal surgery): a multicentre international retrospective cohort study

The laparoscopic approach represents the standard of treatment for renal and adrenal diseases, and its use is increasing even outside referral centres. Although most procedures are routinely performed, intraoperative complications do not occur, and the rate and predictive factors of these complications have not been established. The aim of this study was to evaluate the incidence and type of intraoperative complications and to identify predictive factors in patients undergoing laparoscopic renal and adrenal surgery. This was a cohort, multicentre, international retrospective study. Patients who underwent laparoscopic renal and adrenal surgeries between April 2017 and March 2022 were included in the study. Bivariate analysis was performed using contingency tables and the χ2 test for independent samples to compare qualitative variables and the T test and Mood test for continuous variables. Multivariate analysis was performed using a logistic regression model to obtain adjusted odds ratios. A total of 2374 patients were included in the study. Intraoperative complications were reported for 8.09% of patients who underwent renal surgery, with the most common complications reported being hollow viscus and vascular complications, and for 6.75% of patients who underwent adrenal surgery, with the most common complication reported being parenchymatous viscous complications. Multivariate analysis revealed that both adrenal and renal surgery radiological preoperative factors, such as invasive features during adrenalectomy and the RENAL score during nephrectomy, are predictive factors of intraoperative complications. In contrast to existing data, surgeon experience was not associated with a reduction in the incidence of perioperative complications.

Since the use of the laparoscopic approach for adrenalectomy was initially reported by Gagner et al. in 1992  1 , laparoscopic adrenalectomy (LA) has been the standard of care for the treatment of all benign adrenal masses because it is associated with reduced postoperative pain, early oral intake, and short hospital stays 2 .During the same period, Clayman published the first case series on laparoscopic nephrectomy, reporting the same advantages as the minimally invasive approach 3 .Since then, the use of laparoscopy, including partial nephrectomy (LPN) and nephroureterectomy for urothelial upper tract carcinoma, for accessing retroperitoneal organs has increased rapidly.
Furthermore, with improvements in technology and the use of new surgical techniques, such as the retroperitoneal approach, laparoscopic surgery has been meaningfully improved and increasingly adopted to the extent that the laparoscopic approach to renal surgery, for both radical and partial nephrectomy, is considered the standard procedure at many institutions whenever feasible.Compared with open surgery, laparoscopic surgery has been proven to have identical long-term oncologic outcomes [4][5][6][7] and added benefits, such as shorter hospital stays, lower analgesic requirements 8,9 , and shorter convalescence times.Therefore, despite the widespread use of robotic surgery, the purely laparoscopic approach is still considered the treatment of choice for many benign and malignant diseases, including complex cases for which surgery 10 , such as general and endocrine surgery, as well as urologic surgery (0.7-5.4%) may be difficult.Nevertheless, potentially life-threatening complications during laparoscopic renal and adrenal surgery, including bowel injury (0.8%) 11 , spleen injury (1.4%), pancreatic injury (0.4%) 12 , diaphragmatic injury (0.6%) 13,14 , and vascular complications (0.7-5.4%) are still being reported 15 .Indeed, accessing the retroperitoneal space is a challenge for laparoscopic surgeons because of the need to carefully control veins and arteries that are located deep, behind other structures, and in close proximity to the hollow and parenchymatous viscus.The aim of this study is to determine the rate of intraoperative complications of adrenal and renal surgery by retrospectively examining a large international multicentre database and to identify the predictive factors of perioperative complications.

Methods
This multicentre international retrospective study included patients who underwent laparoscopic renal and adrenal surgery between April 2017 and March 2022.Seven centres were in Italy (Palermo, Roma, Ancona, Napoli, Torino, Genova, Cagliari), two were in Spain (Madrid, Barcelona), one was in Mexico, one was in Argentina, and one was in Romania (i-CLARAS Study Collaborative Group).Only patients treated by the laparoscopic approach were considered because not all participating centres have access to the robotic platform.This clinical study is referred to as the i-CLARAS (intraoperative Complication in Laparoscopic Renal and Adrenal Surgery) study and was publicly registered and approved by the ethics committee of the promoting centre (University Hospital Policlinico of Palermo).All the research was performed in accordance with the relevant guidelines.Informed consent was obtained from all participants and/or their legal guardians.This work has been reported in line with the STROCSS criteria 16 .The inclusion and exclusion criteria are presented in Fig. 1.
Preoperative data.The following preoperative data were retrospectively collected: age, BMI, sex, comorbidities (hypertension and diabetes), and preoperative estimated glomerular filtration rate (GFR).To evaluate patients' general performance status, the Charlson Comorbidity Index was calculated, and the ASA score was collected for each patient.The preoperative bleeding risk was assessed, and the following aetiologies were recorded: previous surgery, direct-acting oral anticoagulants (DAOCs), or haematologic disease.In the renal group, details on the preoperative diagnosis were collected, including the pathological characteristics (benign or malignant disease) and details regarding the pathological diagnosis.Malignant diseases were classified as clear cell renal cell carcinoma (ccRCC), papillary renal cell carcinoma (RCC), chromophobe RCC, Bellini RCC, unclassified, or other.Benign diseases were classified as oncoytoma, renal angiomyolipoma (AML), xanthogranulomatous pyelonephritis, or polycystic kidney/renal cyst.For malignant diseases, the preoperative extent of the primary tumour according to the TNM staging system was previously reported.The following preoperative radiological data were collected: side (right or left), PADUA score, PADUA risk category score, and RENAL score.In the adrenal group, the following preoperative data were collected for each patient: lesion side, number of lesions,

Inclusion criteria Adrenal group
Renal group Radiological and/or laboratory diagnosis of adrenal disease, both functioning and non functioning Radiological and/or laboratory diagnosis of renal disease, both malignant and benign

Age>18 years
Age>18 years Surgery performed through laparoscopic approach Surgery performed through laparoscopic approach Exclusion criteria Open surgery approach Robotic Approach Patients who underwent multivisceral resection and/or tumors of other organs.

Statistical analysis
All the data are presented as the mean ± standard deviation (SD) for continuous variables, median (interquartile range) for ordinal variables, and contingency tables for qualitative variables.Univariate and multivariate analyses were performed to assess the association between preoperative and intraoperative data and the rate of intraoperative complications.Bivariate analysis was performed using the chi-square test for independent samples for qualitative variables, and the t test and Mood test were used for continuous and ordinal variables, respectively.Multivariate analysis was performed, and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated via logistic regression models.A two-tailed, P value < 0.05 indicated statistical significance.Statistical analysis was conducted by a biomedical statistician using R software (Core Team, Vienna, Austria, 2013).

Intraoperative complications
Intraoperative complications were reported for 8.09% (n.123) of patients who underwent renal surgery and for 6.75% (n.27) of patients who underwent adrenal surgery.In the adrenal group, the most frequent complications reported were parenchymatous viscous complications (48.15%), followed by vascular complications (14.81%) and hollow viscus complications (7.41%) (Tables 1 and 2).Particularly, for patients who underwent adrenal surgery and who had invasive features, the rate of intraoperative complications was 25%, and the most frequent complications were parenchymatous viscous complications (3 patients, one patient had splenic injury, and two patients had other parenchymatous viscous lesions); one patient also experienced vascular complications, and no patient experienced hollow viscus complications.In the renal group, the most frequent complications were hollow viscus (39.84%) and vascular (39.3%), followed by parenchymatous viscous complications (25.20%).

Conversion rate
The overall conversion rates were 0.81% (16 patients) in the renal group and 3.42% (14 patients) in the adrenal group; in patients who suffered intraoperative complications, the conversion rates were 13.01% in the renal group and 48.15% in the adrenal group.

Short-term complications
The overall short-term complication rate was 13.45% (55) in the adrenal group and 16.44% (323) in the renal group (Table 1).

Factors predicting intraoperative complications
The results of the univariate analysis are shown in Tables 2 and 3.In the adrenal group, the multivariate analysis revealed that the presence of invasive features was a borderline predictive factor for intraoperative complications (OR 3.57, p = 0.0708).In the renal group, sex, BMI, Charlson Comorbidity Index (CCI) score, surgeon experience, and cTNM were not significant.According to our multivariate analysis, the presence of malignant disease and the use of a retroperitoneal approach were protective factors against intraoperative complications (OR 0.400, p = 0.012 and OR 0.218, p = 0.001, respectively).With regard to patients who underwent partial nephrectomy, according to both univariate and multivariate analyses, a higher RENAL score was associated with a higher incidence of intraoperative complications (OR 1.279, p < 0.001).

Table 1.
Preoperative, intra-operative and post-operative characteristics of patients in the adrenal surgery (n = 409) and Preoperative, intra-operative and post-operative characteristics of patients in the renal group (n = 1965).www.nature.com/scientificreports/
According to the univariate analysis of patients who underwent renal surgery, a higher ASA score was associated with a higher rate of postoperative blood transfusion (p < 0.001), but there were no differences in the rate of intraoperative blood transfusion.www.nature.com/scientificreports/

Factors predictive of conversion
According to the univariate analysis of patients who underwent adrenal surgery, the mass dimension was not significantly associated with conversion; however, the presence of invasive features and low surgeon experience were significantly associated with a higher conversion rate (p = 0.002 and p = 0.019, respectively).According to our multivariate analysis (Table 4), low surgeon experience (OR 5.146, 95% CI 1.072, 24.694; p = 0.041), a longer operation time (OR 1.005, 95% CI 1.001, 1.011; p = 0.048), and the presence of invasive features (OR 14.033, 95% CI 2.606, 75.549; p = 0.002) were significant predictive factors of conversion.In the renal group, the PADUA and RENAL scores were not associated with conversion, whereas a higher cTNM and lower surgeon experience were significantly associated with higher conversion rates (p = 0.001 and p = 0.026, respectively).According to our multivariate analysis (Table 4), low surgeon experience was a borderline risk factor for conversion (OR 4.277, p = 0.051), and a longer operation time (OR 1.012, IC 1.001, 1.019, p < 0.001) and a higher CCI score (OR 1.459, IC 1.061, 2.004, p value 0.020) were significant predictive factors for conversion.

Discussion
Since its introduction at the beginning of the twentieth century, laparoscopic surgery of the retroperitoneal organs has gained interest and enthusiasm.However, in addition to its known advantages, laparoscopy is not free from the risk of intraoperative complications, and their incidence may be underestimated outside major referral centres.Overall, in this study, the rate of intraoperative complications was 6.75% in the adrenal surgery  www.nature.com/scientificreports/group, with the most common complication reported being parenchymatous viscous complications.Apparently, this rate is higher than that reported in a large prospective multicentre study by Bergamini et al., who reported an overall perioperative complication rate of 7.9%, and a intraoperative accident rate of 3.6%.However, considering the rate of intraoperative complications reported in referral and nonreferral centres (2% and 8.2%, respectively), it appears that in nonreferral centres, the rate is similar (6.75% vs. 8.2%), thus highlighting not only the inhomogeneity but also the robustness of the data from this study, which represents the clinical reality worldwide.Furthermore, the rate of complications reported during and immediately after laparoscopic adrenalectomy is inhomogeneous; some authors have reported a rate of 4.9% 17 , whereas others 18 have reported a rate of intraoperative complications of 33.3%.This discrepancy reflects the heterogeneity of studies that included adrenalectomy for pheocromocytoma, for which higher rates of perioperative complications are reported, mostly hypertensive crisis, and surgery for other indications.Surprisingly, this study showed that the most common intraoperative complication was parenchymatous viscous complications, with vascular complications occurring at a lower rate.These results contrast with the results of a review by Strebel et al. 15 , who reported that the most common complication of laparoscopic adrenal surgery was vascular injury 5 .This finding was probably due to a greater focus on potentially serious vascular complications than on parenchymatous visceral injuries.The results obtained in this study, on the other hand, showed that in routine surgical practice, the most frequent injuries were to parenchymatous organs, which often have no counterpart in terms of worsening postoperative outcomes and are therefore generally not considered or misrecognized in the literature.According to a multi-institutional retrospective study conducted in 2011, known risk factors for the occurrence of intraoperative complications in laparoscopic adrenal surgery include low surgeon experience, pheochromocytoma, age, BMI, and mass dimensions 19 .Nevertheless, whether mass size is a risk factor for poor surgical outcomes of laparoscopic surgery is unclear.
Bergamini et al. 17 and Shen et al. 20 reported that a larger adrenal mass influences laparoscopic outcomes, whereas other authors did not [21][22][23] .The results of this study suggest that the presence of imaging features indicating local invasion (infiltration of the surrounding structures, venous invasion, and absence of an adipose cleavage plane), rather than the mass dimension 20 , could be a predictive factor of intraoperative complications and conversion, thus underlining the importance of preoperative radiological assessment to accurately clarify the morphological characteristics of adrenal masses for planning surgery.A second pivotal theme regards the role of surgical experience as a protective factor against complications, as some authors have reported that surgical volume is a predictor of better outcomes in patients undergoing adrenalectomy 24 .In this study, no significant relationship was found between surgical experience and complications.However, in line with the findings of other authors 1,15,25,26 , surgical experience was a predictive factor for conversion (p = 0.019).In this multicentre study, the overall shortterm complication rate in patients who underwent adrenal surgery was 13.45%.This rate is slightly higher than that reported in the literature 17 .Notably, the rate of short-term complications was significantly higher than the rate of intraoperative complications, and most published studies have focused on postoperative outcomes to establish the safety of laparoscopic surgery; however, these results emphasize that postoperative outcomes alone could mask the real advantages and disadvantages of the laparoscopic approach.In the renal group, the rate of intraoperative complications was 8.09%, and the most common complications reported were hollow viscous and vascular complications.Overall, the rate of complications reported was similar to that previously reported in monocentric retrospective studies, suggesting that this could represent the average complication rate for renal laparoscopic surgery.Apparently, the proportion of hollow viscus complications was greater than that reported in other single-centre studies 11,27 .However, the majority of hollow viscous complications involved upper urinary tract structures (the calyx and renal pelvis), whereas bowel injuries were reported in only ten patients, reflecting a low rate of gastrointestinal injury.Surprisingly, although most nephrectomies were performed by urologists, the most common hollow viscous involved during partial nephrectomy was the calyx and renal pelvis.In this study, the rate of vascular complications was similar to the rate of hollow viscous complications (39.3%).Of these, 24 were classified as minor vascular complications and were managed through sutures or open conversion, and 26 were classified as major vascular complications and were managed mainly through open conversion.In contrast to the findings of other studies 28 , preoperative factors such as age and ASA score did not result in a higher complication rate.In contrast to the findings reported by other authors 29 , the rates of vascular and hollow viscous complications were comparable for patients treated via the transperitoneal approach.In this study, the retroperitoneal approach was found to be a protective factor against intraoperative complications, in line with the literature.It is conceivable that the retroperitoneal approach could be safer for patients who previously underwent abdominal surgery because it avoids the need for adhesion lysis.Another explanation for the protective role could be that surgeons who use the retroperitoneal approach have more experienced than those who use only transperitoneal access or who treat less difficult cases.In patients who underwent partial nephrectomy, a higher RENAL score was associated with a higher incidence of intraoperative complications (OR 1.279, p < 0.001), indicating the difficulty of these surgical procedures.With regard to the pathological characteristics of the disease, in the univariate analysis, malignant disease was associated with a higher incidence of intraoperative complications, whereas in the multivariate analysis, it was a protective factor.It is conceivable that the results of the univariate analysis did not take into account the role of confounding variables, and the statistical analysis showed that patients with malignant disease who experienced intraoperative complications had a significantly higher RENAL score than patients with benign disease who experienced intraoperative complications.Moreover, the percentage of procedures performed by young surgeons for patients with benign disease who experienced intraoperative complications was not significantly higher than the percentage of procedures performed by senior surgeons for patients with malignant diseases; therefore, neither surgeon experience could explain this result.In contrast, it seems that the type of surgery performed could have influenced the outcome in patients with benign disease since the proportion of patients who underwent partial nephrectomy was significantly higher in patients with benign disease who experienced intraoperative complications than in patients with malignant disease who did not experience intraoperative complications (p = 0.005  .319,3.372, p = 0.002) were significant predictive factors of short-term postoperative complications, whereas radical nephrectomy was a protective factor for short-term postoperative complications (OR 0.697, 95% CI 0.499, 0.973, p = 0.034).These results are not surprising considering the greater technicaldifficulties associated with nephroureterectomy than with radical nephrectomy.With regard to the hand-assisted laparoscopic approach, we know that this approach is generally used by surgeons with the aim of providing better control of potential vascular complications (i.e., massive bleeding from the main vessels) rather than acting as a "retractor" into the abdominal cavity.Furthermore, compared with pure laparoscopy, the handassisted technique might provide faster organ removal (i.e., kidney extraction) once the procedure has been finalized.Despite the theoretical advantages of the hand-assisted approach, it is no longer adopted by more skilled laparoscopic surgeons who prefer pure conventional laparoscopy, even for the most challenging cases.Indeed, hand-assisted surgery is mostly performed by colleagues with less experience as a sort of "safer approach"; however, limited surgical experience can often result in more intra-and postoperative complications (an increased overall complication rate).On the other hand, in this study, surgeon experience did not affect the results for the following reasons.In all the centres included in our analysis (i.e., academic hospitals), when junior colleagues and/or less skilled surgeons performed the procedures (especially those that were technically demanding and difficult), they worked alongside more experienced surgeons to avoid and, if needed, properly manage complications.As a result, we minimized any risk for the patients, and did not therefore significantly affect the outcomes during the process of mastering the surgical learning curve.Finally, with regard to factors influencing the need for perioperative blood transfusions, the results of this study suggest that the postoperative blood transfusion rates are not related to the type of surgical procedure but instead depend on patient preoperative risks and comorbidities.This study has several limitations.First, the same surgical region (retroperitoneal space) was used for each of the different surgeries.Second, a subgroup analysis for benign and malignant diseases was not performed, even though malignant masses are more challenging to access than benign masses.Finally, it could be argued that the study being a retrospective multicentre study lowered the weight of the results.However, the retrospective design is not a limitation, as it avoids distortion of clinical reality and surgeons' reticence in reporting intraoperative complications.Despite these limitations, this study provides valuable insights into the possible disadvantages of laparoscopic surgery for tumours involving the adrenal gland and kidney.The results of this study can be translated to clinical practice, even in centres that not considered high-volume centres, and can be used to identify tumour-related factors that can predict intraoperative complications.

Conclusion
The results of this multicentre international study showed overall intraoperative complication rates of 6.75% for laparoscopic adrenal surgery and 8.09% for laparoscopic renal surgery.The multivariate analysis revealed that radiological preoperative factors, such as invasive features for patients undergoing adrenalectomy and the RENAL score for those undergoing nephrectomy, are predictive factors of intraoperative complications.In contrast to existing data, surgeon experience was not associated with a reduction in the incidence of perioperative complications.

Table 2 .
Univariate analysis for factors associated with intraoperative and short term post-operative complications in the adrenal surgery group (n = 409).Significant values are in bold.

Table 3 .
Univariate analysis for factors associated with intraoperative and short term post-operative complications in the renal surgery group(n = 1965).Significant values are in bold.