Predictors of Medical and Surgical Complications After Robot-Assisted Partial Nephrectomy: An Analysis of 1139 Patients in a Multi-Institutional Kidney Cancer Database

Abstract

Introduction:

Previous robot-assisted partial nephrectomy (RAPN) studies have identified various predictors of overall and major postoperative complications, but few have evaluated the specific role of these factors in the development of medical and surgical complications. In this study, we present an analysis of the modifiable and nonmodifiable variables influencing medical and surgical complications in a contemporary series of patients who underwent RAPN and were followed in a prospectively maintained, multi-institutional kidney cancer database.

Methods:

A retrospective review of all patients who underwent RAPN at four institutions between 2008 and 2015 was performed. Multivariable logistic regression models were used to determine predictors of medical and surgical postoperative complications.

Results:

Data from 1139 patients were available for analysis. Sixty-seven patients (5.8%) experienced a medical postoperative complication, and 82 (7.1%) experienced a surgical complication. Decreasing baseline estimated glomerular filtration rate (eGFR) (odds ratio [OR] = 0.98, p = 0.003), greater estimated blood loss (EBL) (OR = 1.002, p = 0.001), and operating surgeon (OR = 8.01, p < 0.001) were associated with an increased likelihood of surgical complications, while decreasing baseline eGFR (OR = 0.99, p = 0.054) and operating surgeon (OR = 1.96, p = 0.054) were associated with an increased likelihood of medical complications.

Conclusion:

We present complication risks in a large contemporary cohort of patients undergoing robotic partial nephrectomy (RPN) with only 11.3% of patients experiencing a medical or surgical postoperative complication. Prospective candidates for robotic PN with poor baseline renal function and/or risk factors for greater EBL, including a high body mass index, or a complex renal mass should be counseled appropriately on their increased risk for a medical or surgical postoperative complication.

Introduction

Since its 2009 guideline for management of the clinical T1 renal mass, the American Urological Association has considered nephron-sparing approaches in all patients with a clinical T1 renal mass to be a standard of care.¹ Robot-assisted partial nephrectomy (RAPN) has gained traction since its introduction in 2004 as an alternative to open and traditional laparoscopic PN due to its equivalent oncological and functional outcomes, while both improving patient convalescence and shortening a surgeon's learning curve.² In one of the largest series to date, RAPN has also been shown to achieve a low complication rate of 14.4%,² which is lower than historical complication rates of open and laparoscopic PN, reported to be 19.2% and 24.9%, respectively, in one of the largest open and laparoscopic PN series to date.³

Prior PN studies have identified various predictors of overall and major postoperative complications,^4

–8 and the role of these factors has been found to differ across studies. However, few studies have evaluated the role of these factors in the development of surgical complications, those which are related to the performance of the procedure (incisional complication, unplanned event during surgery), and medical complications, those where a patient develops a medical condition following surgery, or where a patient's preexisting medical condition is worsened by perioperative stress.⁹ By identifying factors that increase the risk of medical and surgical complications, patients could be better counseled regarding their risk for complications, and modifications could be implemented to limit them going forward.

We therefore present an analysis of the variables influencing medical and surgical complications in a contemporary series of 1139 patients who underwent RAPN at a total of four medical centers, and were followed in a prospectively maintained, multi-institutional kidney cancer database.

Methods

This retrospective study relied on an institutional review board-approved multi-institutional kidney cancer database to identify 1145 patients undergoing RAPN at four medical institutions between January 2008 and June 2015. All data were managed and prospectively maintained in Research Electronic Data Capture (REDCap).¹⁰ Patients with missing postoperative complication data (n = 6) were excluded from analysis. Of the 1139 included patients, demographic, clinical, and operative data, including age, gender, body mass index (BMI), American Society of Anesthesiologists (ASA) score, history of prior abdominal surgery, baseline estimated glomerular filtration rate (eGFR), serum creatinine, RENAL nephrometry score, tumor location and size, arterial clamping technique, collecting system entry, warm ischemia time (WIT), operative time, and estimated blood loss (EBL), were downloaded from REDCap for analysis.

The Charlson comorbidity index (CCI), which assigns distinct weights to comorbidities (e.g., hypertension (HTN) = 1, leukemia = 2), was used to categorize comorbidities for all patients.¹¹ Patients with cirrhosis or patients lacking sufficient data to differentiate between mild or severe liver disease were defaulted as mild and given a weight of 1. Since patients with aneurysms only ≥6 cm are counted as having peripheral vascular disease (PVD), patients with a history of an aneurysm with data lacking on the size of the aneurysm were not assigned a weight of 1 for PVD. Patients with diabetes and any concomitant end-organ disease (i.e., neuropathy, retinopathy, nephropathy) were given a weight of 2.

Postoperative complications occurring within the first 30 days after surgery were categorized according to the Clavien–Dindo Classification System,¹² and postoperative complications with a Clavien score ≥3 were classified as major. Postoperative complications were also classified as surgical or medical according to the previously defined method discussed above.⁹

Statistics

For all demographic, clinical, intraoperative, and postoperative variables, medians and interquartile ranges were calculated for continuous variables, and frequencies and percentages were calculated for all categorical variables. Rates for overall, major, medical, and surgical postoperative complications were calculated for all patients.

Univariable and multivariable logistic regression models, including age, gender, BMI, baseline eGFR, tumor size, RENAL score, CCI, EBL, WIT, surgeon, and date of surgery, were used to determine which variables were associated with postoperative complications (i.e., overall, major, medical, surgical). The influence of surgeon experience defined as consecutive robotic partial nephrectomy (RPN) number of the surgeon was also assessed in multivariable analysis. Multivariable regression analysis was also used to assess the influence of these variables on EBL.

Results

Patient characteristics and perioperative outcomes

All patient characteristics and perioperative outcomes are presented in Table 1. For the 1139 patients, median age, BMI, tumor size, and RENAL score were 60 years old, 29.5 kg/m², 3.0 cm, and 7, respectively. One-hundred seventy-seven (15.5%) patients had baseline chronic kidney disease (baseline eGFR <60 mL/min/1.73 m²). Median WIT and EBL were 15 minutes and 100 mL, respectively. Median CCI was 2.0.

Table 1.

Patient Characteristics and Perioperative Outcomes

Patients	1139
Date of surgery	2012 (2011 to 2013)
Age (years)	60.0 (51.0 to 68.0)
Male	651 (57.2%)
BMI	29.5 (25.9 to 34.1)
ASA score	3.0 (2.0 to 3.0)
CCI	2.0 (1.0 to 3.0)
Prior abdominal surgery	561 (49.3%)
Serum creatinine (mg/dL)	0.90 (0.77 to 1.10)
Baseline eGFR (mL/min/1.73 m²)	82.6 (67.8 to 99.6)
CKD stage 1	440 (39.5%)
CKD stage 2	598 (44.7%)
CKD stage 3	162 (14.5%)
CKD stage 4	13 (1.2%)
CKD stage 5	2 (0.2%)
Tumor size (cm)	3.0 (2.1 to 4.0)
RENAL score	7.0 (6.0 to 9.0)
Anterior tumor location	370 (37.0%)
Renal artery clamp technique
Main artery clamping	882 (77.4%)
Other^a	257 (22.6%)
Collecting system entry	366 (50.3%)
WIT (minutes)	15.0 (11.0 to 20.0)
EBL (mL)	100.0 (50.0 to 150.0)
Operative time (minutes)	183.0 (151.0 to 224.0)
First 30 postoperative days
% change eGFR	−13.6% (−26.2% to 0.0%)
AKI	234 (27.1%)
12 months postsurgery
% Change eGFR	−11.0% (−23.0% to 0.0%)
Progression to CKD	91 (13.4%)
Length of stay (days)	1.0 (1.0 to 2.0)
NNPV removed (cm³)	11.1 (5.3 to 25.6)
Pathologic stage > T1	61 (7.3%)
PSM (patients)	50 (5.2%)

For categorical variables, frequencies presented with percentages in parentheses.

For continuous variables, medians presented with interquartile ranges in parentheses.

Includes selective arterial clamping, off-clamp, and microvessel clamping.

AKI = acute kidney injury; ASA = American Society of Anesthesiologists; BMI = body mass index; CCI = Charlson comorbidity index; CKD = chronic kidney disease; EBL = estimated blood loss; eGFR = estimated glomerular filtration rates; NNPV = nonneoplastic parenchymal volume; PSM = positive surgical margins; WIT = warm ischemia time.

Complications

Overall, 158 postoperative complications were identified in 130 patients (11.3%). Of these, 43 major, 71 medical, and 86 surgical complications were identified in 42 (3.7%), 67 (5.8%), and 82 (7.1%) patients, respectively (Table 2).

Table 2.

Postoperative Complication Data

Overall complications
Number of patients	130 (11.3%)
Total	158
Clavien grade
I	46 (4.0%)
II	69 (6.0%)
III	27 (2.4%)
IV	15 (1.3%)
V	1 (0.09%)
Major complications
Number of patients	42 (3.7%)
Total	43
Medical complications
Number of patients	67 (5.8%)
Total	71
Surgical complications
Number of patients	82 (7.1%)
Total	86

Results from multivariable analyses assessing the influence of the demographic, clinical and perioperative variables on overall, major, medical, and surgical postoperative complications are presented in Table 3. Decreasing baseline eGFR (odds ratio [OR] = 0.99, p = 0.054) and operating surgeon (OR = 1.96, p = 0.054) were associated with an increased likelihood of medical complications. Increasing EBL (OR = 1.002, p = 0.001) and decreasing baseline eGFR (OR = 0.98, p = 0.003) and operating surgeon (OR = 8.01, p < 0.001) were associated with an increased likelihood of surgical complications. Surgeon experience was not associated with surgical complications (OR = 0.99, p = 0.320) or medical complications (OR = 1.00, p = 0.990).

Table 3.

Multivariable Logistic Regression Analyses Predicting Postoperative Complications

	Overall complications		Major complications		Medical complications		Surgical complications
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Age	0.99 (0.96, 1.02)	0.483	0.97 (0.93, 1.01)	0.144	0.99 (0.96, 1.03)	0.729	0.99 (0.95, 1.02)	0.487
Gender	1.07 (0.69, 1.66)	0.762	0.93 (0.51, 1.97)	0.994	1.28 (0.71, 2.31)	0.414	1.15 (0.66, 2.01)	0.631
BMI	1.01 (0.98, 1.05)	0.513	1.02 (0.96, 1.07)	0.590	1.03 (0.99, 1.08)	0.154	1.00 (0.96, 1.05)	0.859
CCI	1.03 (0.85, 1.25)	0.734	1.20 (0.91, 1.59)	0.200	1.09 (0.85, 1.38)	0.503	0.96 (0.74, 1.24)	0.729
Baseline eGFR	0.98 (0.98, 0.99)	0.002	0.98 (0.97, 0.99)	0.035	0.99 (0.98, 1.00)	0.054	0.98 (0.97, 0.99)	0.003
Tumor size	1.08 (0.94, 1.24)	0.298	1.04 (0.84, 1.31)	0.702	1.01 (0.82, 1.25)	0.923	1.13 (0.96, 1.33)	0.151
RENAL score	0.96 (0.85, 1.04)	0.556	0.99 (0.81, 1.21)	0.923	0.89 (0.75, 1.05)	0.170	1.06 (0.89, 1.24)	0.516
WIT	1.01 (0.99, 1.04)	0.371	1.02 (0.98, 1.06)	0.429	1.02 (0.99, 1.00)	0.275	1.00 (0.97, 1.03)	0.933
EBL	1.00 (1.00, 1.00)	0.005	1.00 (1.00, 1.00)	0.009	1.00 (0.99, 1.00)	0.804	1.00 (1.00, 1.00)	0.001
Date of surgery	1.00 (1.00, 1.00)	0.519	1.00 (1.00, 1.00)	0.391	1.00 (1.00, 1.00)	0.945	1.00 (1.00, 1.00)	0.251
Operating surgeon	3.63 (2.08, 6.33)	<0.001	5.32 (1.92, 14.75)	0.001	1.96 (0.99, 3.87)	0.054	8.01 (2.14, 16.9)	<0.001

CI = confidence interval; OR = odds ratio.

Bold p-values are significant at <0.05 level.

For all the surgeons, surgeon experience did not significantly reduce the incidence of medical complications (p = 0.067, p = 0.466, p = 0.810, p = 0.421) or surgical complications (p = 0.378; p = 0.240; p > 0.999; p = 0.281).

Separate multivariable models assessing the role of EBL and baseline eGFR on complications were also used. In one surgeon, only one surgical complication occurred and a separate analysis for surgical complications was not conducted for this surgeon. For all other surgeons, EBL (p < 0.001, p = 0.014, p = 0.030) and lower baseline eGFR (p = 0.001, p = 0.022, p = 0.034) were associated with an increased risk of postoperative surgical complications. For only one surgeon, decreasing baseline eGFR increased the likelihood of a medical complication (p = 0.028, p = 0.179, p = 0.259, p = 0.846).

Estimated blood loss

BMI (β = 0.00, p = 0.022), prior abdominal surgery (β = 0.08, p = 0.025), tumor size (β = 0.13, p = 0.001) and RENAL score (β = 0.11, p = 0.003) were associated with greater EBL, while main renal artery clamping (β = −0.16, p < 0.001) and female vs male (β = −2.64, p = 0.009) were associated with reduced EBL. Increasing surgeon experience was also associated with less EBL (β = −0.14, p = 0.019).

Discussion

This retrospective review of a prospectively maintained, multi-institutional kidney cancer database presents to our knowledge, the largest series of rates of and factors associated with postoperative medical and surgical complications after RAPN. Of the 158 overall complications that occurred in 11.3% of our cohort, more patients had a surgical (7.1%) compared to a medical (5.8%) postoperative complication.

Factors associated with an increased risk of developing a medical condition following surgery, or exacerbating a preexisting medical condition, were decreasing baseline eGFR. Interestingly, increasing CCI was not associated with a significantly increased risk of medical complications (OR = 1.09, p = 0.503). This suggests that the presence of more severe comorbidities does not increase the likelihood that these comorbidities will worsen after surgery. That decreasing baseline eGFR increases risk of medical complication highlights the need for the implementation of renal function preservation techniques in appropriately selected candidates. Cold ischemia when extended ischemia is expected or off-clamp technique in addition to zero-ischemia technique with superclamping are techniques proposed that potentially preserve renal function compared to hilar clamping with warm ischemia.¹³

Variables associated with a significantly increased risk of surgical complications were decreasing baseline eGFR, operating surgeon, and increasing EBL. The adverse effect of a lower baseline eGFR and increasing EBL was not only observed overall but was also observed individually for each surgeon. While grading systems for renal tumors such as the RENAL nephrometry score¹⁴ and preoperative aspects and dimensions used for anatomical (PADUA) classification¹⁵ have been validated as independent predictors of complications following PN, with four times greater odds of complications when the RENAL score is 9 or greater and when the PADUA classification is 10 or greater,¹⁶ our multivariable analysis has shown that tumor size and RENAL score were not significant predictors of overall, major, medical, or surgical complications. This finding objectively defines the utility afforded by our prospectively maintained, multi-institutional database, and likely would have otherwise been missed in a less robust review from a single center. Similarly, Tanagho and colleagues reported that the correlation of RENAL score with complications following RAPN was small (OR = 1.18),¹⁷ and PADUA classification has been found to be a significant, but weak predictor of complications following RAPN (OR = 1.04).⁸ This decreased strength of correlation between anatomic renal tumor grading systems and postoperative complications after RAPN as opposed to laparoscopic or open PN may also be a result of increased surgeon experience and the technical advantage afforded by robot-assisted laparoscopic surgery.¹⁸ For example, a recent retrospective analysis noted that surgeon experience was an independent predictor of WIT >20 minutes (OR = 6.381, p < 0.001) and overall complication rates (OR = 3.937, p < 0.001) after RAPN.⁸

While our multivariable analysis indicates that tumor size and RENAL score are not significant predictors of surgical complications, it does demonstrate that tumor size and RENAL score, along with increasing BMI and prior abdominal surgery, are associated with greater EBL. Because increasing EBL has been shown to be associated with an increased risk of surgical complications, it is likely that tumor size and RENAL score, which are primary determinants of EBL, underlie why EBL is associated with an increased risk of postoperative surgical complications. As far as the other factors associated with increasing EBL, the relationship with prior abdominal surgery possibly reflects the need to perform an extensive lysis of adhesions.

In addition, while surgeon experience overall and for each surgeon did not reduce complication risk, the association of increased surgeon experience with decreased blood loss highlights that high- vs low-volume surgeon can potentially reduce blood loss and subsequently reduce the risk of a patient experiencing a surgical complication. Also, it has been shown that uninterrupted suturing with tightening and locking after stitching compared to interrupted suturing¹⁹ reduces blood loss and can potentially reduce the risk of surgical complications.

The limitations of this study are comparable to those of other large series of RAPN. Specifically, referral of more complex renal masses to the four institutions contributing to the multi-institutional database introduces a selection/referral bias that may also affect surgical outcomes.¹⁷ Also, it should be noted that trainees were involved in RPNs as console surgeons and their level of involvement was based on PGY level. However, the lack of case by case data on trainee involvement inhibits our ability to comment on the effect of trainee involvement on the risk of postoperative complications. Furthermore, the quality of data in our retrospective review of prospectively maintained databases is significantly higher compared with a purely retrospective study, but does fall short of what would be derived from a truly prospective study.¹⁷

Strengths of this study are higher internal and external validity of this analysis is highly relative to single-center reports. Having adjusted for operating surgeon in our analysis, the confounding influence of intersurgeon variability is controlled for; as a result, the influence of each factor included in the multivariable models is more accurately represented than in a single surgeon report. In addition, this multi-center report is more generalizable to the United States population than a single surgeon report since this report represents four as opposed to one center. As evidenced in this study, the influence of decreasing baseline eGFR on medical complications was inconsistent across surgeons, increasing the risk of medical complications for only one surgeon, but not the other three; however, the collective analysis of all surgeons highlighted that decreasing baseline eGFR does indeed increase the risk of medical complications.

Also, a collaboration of institutions allowed for a robust analysis of relatively low occurring outcomes (86 surgical [7.1%] and 72 medical complications [5.8%]), which if studied at the single-surgeon level, would have yielded a low number of events and a subsequent weak analysis. Indeed, for one surgeon in particular, the rate of surgical complications was 0.7%. Ultimately, this retrospective review of a prospectively maintained, multi-institutional kidney cancer database presents, to our knowledge, the largest series of the rates of and factors associated with postoperative overall, major, medical, and surgical complications after RAPN. The identification of those modifiable and nonmodifiable risk factors for the exacerbation of medical and surgical postoperative complications will enable patients to be better informed regarding their risk for complications. For example, those patients with compromised renal function who have preexisting medical conditions may be at a higher risk for exacerbation of these conditions postoperatively and should therefore be counseled accordingly. Patient–provider discussions such as these should ultimately serve to facilitate the choice of the most appropriate treatment modality for a patient with a small renal mass.

For example, when counseling a patient regarding his/her treatment options for a renal mass, the presence of a higher baseline eGFR should particularly aid in advising a patient of his/her decreased risk for both medical and surgical complications. In addition, while EBL, a predictor of surgical complications can only be determined intraoperatively, there are several patient and radiographic characteristics associated with EBL that should therefore be addressed preoperatively. With regard to EBL, these include BMI, prior surgery, tumor size, and RENAL score. Patient–provider discussions that address these criteria should ultimately serve to facilitate the choice of the most appropriate treatment modality for a patient with a small renal mass.

Conclusion

In this large, contemporary series of postoperative complications following RPN, we report a medical complication rate of 5.8%, a surgical complication rate of 7.1%, and an overall complication rate of 11.3%, which is lower than reported historical complication rates of open and laparoscopic PN. Predictors of medical complications included decreasing baseline kidney function and operating surgeon, and predictors of surgical complications included decreasing baseline kidney function, operating surgeon and higher EBL.

Techniques to minimize ischemic injury in addition to performance of the RPN by a high volume surgeon can potentially lead to preservation of renal function, less blood loss, and ultimately a decreased incidence of medical and surgical postoperative complications. Candidates for RAPN with poor baseline renal function and risk factors for high EBL (i.e., high BMI, complex tumor) should be counseled appropriately on their increased risk for medical and surgical complications.

Footnotes

Author Disclosure Statement

Dr. Eun reports personal fees from Intuitive Surgical, personal fees from Medtronic, and personal fees from Conmed, outside the submitted work. Dr. Abaza reports grants from Intuitive Surgical, grants from ConMed (Surgiquest), outside the submitted work. All other authors have nothing to disclose.

Abbreviations Used

References

Campbell

, Novick

, Belldegrun

, et al. Guideline for management of the clinical T1 renal mass. J Urol, 2009; 182:1271–1279.

Spana

, Haber

, Dulabon

, et al. Complications after robotic partial nephrectomy at centers of excellence: Multi-institutional analysis of 450 cases. J Urol, 2011; 186:417–421.

Gill

, Kavoussi

, Lane

, et al. Comparison of 1,800 laparoscopic and open partial nephrectomies for single renal tumors. J Urol, 2007; 178:41–46.

Larcher

, Fossati

, Tian

, et al. Prediction of complications following partial nephrectomy: Implications for ablative techniques candidates. Eur Urol, 2016:69:676–682.

Mathieu

, Verhoest

, Droupy

, et al. Predictive factors of complications after robot-assisted laparoscopic partial nephrectomy: A retrospective multicentre study. BJU Int, 2013; 112:E283–E289.

Gorin

, Mullins

, Pierorazio

, Jayram

, Allaf

. Increased intra-abdominal fat predicts perioperative complications following minimally invasive partial nephrectomy. Urology, 2013; 81:1225–1230.

Sun

, Bianchi

, Trinh

, et al. Hospital volume is a determinant of postoperative complications, blood transfusion and length of stay after radical or partial nephrectomy. J Urol, 2012; 187:405–410.

Ficarra

, Bhayani

, Porter

, et al. Predictors of warm ischemia time and perioperative complications in a multicenter, international series of robot-assisted partial nephrectomy. Eur Urol, 2012; 61:395–402.

Krane

, Manny

, Mufarrij

, Hemal

. Does experience in creating a robot-assisted partial nephrectomy (RAPN) programme in an academic centre impact outcomes or complication rate?. BJU Int, 2013; 112:207–215.

10.

Harris

, Taylor

, Thielke

, Payne

, Gonzalez

, Conde

. Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform, 2009; 42:377–381.

11.

Charlson

, Charlson

, Peterson

, Marinopoulos

, Briggs

, Hollenberg

. The Charlson comorbidity index is adapted to predict costs of chronic disease in primary care patients. J Clin Epidemiol, 2008; 61:1234–1240.

12.

Clavien

, Barkun

, de Oliveira

, et al. The Clavien-Dindo classification of surgical complications five-year experience. Ann Surg, 2009; 250:187–196.

13.

Volpe

, Blute

, Ficarra

, et al. Renal ischemia and function after partial nephrectomy: A collaborative review of the literature. Eur Urol, 2015; 68:61–74.

14.

Kutikov

, Uzzo

. The R.E.N.A.L. nephrometry score: A comprehensive standardized system for quantitating renal tumor size, location and depth. J Urol, 2009; 182:844–853.

15.

Ficarra

, Novara

, Secco

, et al. Preoperative aspects and dimensions used for an anatomical (PADUA) classification of renal tumours in patients who are candidates for nephron-sparing surgery. Eur Urol, 2009; 56:786–793.

16.

Hew

, Baseskioglu

, Barwari

, et al. Critical appraisal of the PADUA classification and assessment of the R.E.N.A.L. nephrometry score in patients undergoing partial nephrectomy. J Urol, 2011; 186:42–46.

17.

Tanagho

, Kaouk

, Allaf

, et al. Perioperative complications of robot-assisted partial nephrectomy: Analysis of 886 patients at 5 United States centers. Urology, 2013; 81:573–579.

18.

Kim

, Larson

, Figenshau

. Perioperative complications of robot-assisted partial nephrectomy. Curr Urol Rep, 2014; 15:377.

19.

Kawa

, Kinoshita

, Komai

, Inoue

, Masuda

, Matsuda

. Uninterrupted suturing of renal parenchyma in laparoscopic partial nephrectomy decreases renal ischemic time and intraoperative blood loss. Int J Urol, 2010; 17:382–384.