Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266

Original article

Predictors of 30-day acute kidney injury following radical and partial nephrectomy for renal cell carcinoma Marianne Schmid, M.D.a,b,c,*, Abd-El-Rahman Abd-El-Barr, M.D.d, Giorgio Gandaglia, M.D.e, Akshay Sood, M.D.d, Kola Olugbade Jr., M.D.a,b, Nedim Ruhotina, M.D.a,b, Jesse D. Sammon, D.O.d, Briony Varda, M.D.a,b, Steven L. Chang, M.D.a,b, Adam S. Kibel, M.D.a,b, Felix K. Chun, M.D.c, Mani Menon, M.D.d, Margit Fisch, M.D.c, Quoc-Dien Trinh, M.D.a,b a

Center for Surgery and Public Health, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA b Division of Urologic Surgery, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA c Department of Urology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany d Vattikuti Urology Institute, Henry Ford Health System, Detroit, MI, USA e Department of Urology, University Cita-Salute San Raffaele, Milan, Italy Received 28 January 2014; received in revised form 29 April 2014; accepted 3 May 2014

Abstract Introduction: Patients with renal cell carcinoma who were treated with radical nephrectomy (RN) or partial nephrectomy (PN) are at risk of postoperative acute kidney injury (AKI), and in consequence, short- and long-term adverse outcomes. We sought to identify independent predictors of 30-day AKI in patients undergoing RN or PN. Materials and methods: Between 2005 and 2011, patients who underwent RN or PN for renal cell carcinoma within the National Surgical Quality Improvement Program data set were identified. Patients with preexisting severe renal failure, defined as a preoperative estimated glomerular filtration rate o30 ml/min/1.73 m2, were excluded from the analyses. AKI was defined as an elevation of serum creatinine 42 mg/dl above baseline or the need for dialysis within 30 days of surgery. Univariable and multivariable logistic regression analyses were used to examine the association between preoperative factors and the risk of postoperative AKI. Results: Overall, 1,944 (58.6%) and 1,376 (41.4%) patients underwent RN and PN, respectively. Overall, 1.8% of the patients included in the study experienced AKI within an average of 5.4 days after RN or PN. Independent predictors for AKI included obesity (odds ratio [OR] ¼ 2.24, P ¼ 0.04), history of neurovascular disease (OR ¼ 5.29, P o 0.001), and a preoperative chronic kidney disease stage II (OR ¼ 10.00, P ¼ 0.03) or stage III (OR ¼ 26.49, P ¼ 0.02). Furthermore, RN (OR ¼ 2.87, P ¼ 0.02) or the open approach (OR ¼ 2.18, P ¼ 0.04) was significantly associated with postoperative AKI. AKI was significantly associated with adverse postoperative outcomes, such as prolonged length of stay, occurrence of any complication, and mortality (all P o 0.001). Conclusions: The assessment of preoperative kidney function and comorbidity status is essential to identify patients at risk of postoperative AKI. In addition to preoperative chronic kidney disease stages II and III, neurovascular disease, obesity, and surgical approach (RN or open) represent predictors of 30-day AKI. Careful patient selection as well as preoperative planning may help reduce this unfavorable postoperative outcome. r 2014 Elsevier Inc. All rights reserved.

Keywords: Radical and partial nephrectomy; Renal cell carcinoma; Acute kidney injury; Predictors; Outcomes

1. Introduction

* Corresponding author. Tel.: þ1-617-732-6325; þ1-857-210-7801 (mobile); fax: þ1-617-566-3475. E-mail address: [email protected] (M. Schmid).

http://dx.doi.org/10.1016/j.urolonc.2014.05.002 1078-1439/r 2014 Elsevier Inc. All rights reserved.

Renal cell carcinoma (RCC) is among the most common cancers in both men and women, with an estimated 65,150 new cases and 13,680 deaths for the year 2013 in the United States [1]. The gold standard treatment for clinically

1260

M. Schmid et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266

localized RCC is surgical excision by radical nephrectomy (RN) or partial nephrectomy (PN) [2]. Although RN and PN are associated with better cancer control relative to other treatment modalities [3,4], they are associated with a risk of short- and long-term renal injury [5,6]. Indeed, renal dysfunction is a commonly observed adverse event in patients undergoing RN or PN, with some studies showing up to 60% of patients developing some form of kidney dysfunction after surgery [7,8]. This is particularly true for older and sicker individuals, who often present with preexisting chronic kidney disease (CKD) at the time of surgery [9]. However, the incidence of acute kidney injury (AKI) and its effect on adverse outcomes in patients with RCC is not yet sufficiently examined. Clinical findings associating AKI episodes with progressive kidney disease mostly derive from cardiac surgery cohorts or critically ill patients [10,11]. Consequently, effort has to be made to investigate this adverse outcome in the urological surgery setting, with regard to better understanding and identification of AKI determinants. Renal damage following renal surgery can be minimized with appropriate patient and procedure selection, as well as careful operative technique and adequate perioperative resuscitation, as suggested in previous studies [12–14]. Furthermore, previous investigations on the effect of RN or PN on kidney function reflect care at high-volume tertiary referral centers and may not be generalizable to the US population. On the basis of these considerations, we sought to examine the incidence and predictors of postoperative AKI in a large cohort of patients with RCC treated with RN or PN, who were included in the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) [15]. The NSQIP is a prospectively collected database specifically developed to facilitate the assessment of surgical outcomes and complications [16,17]. Though limited by the lack of data on surgical and imaging characteristics, the NSQIP tracks preoperative serum creatinine (SCr) values and specifically measures postoperative progressive renal dysfunction and acute renal failure (ARF). It is noteworthy that the NSQIP does not provide access to a postoperative SCr value for each patient, but only an indication of whether they experienced AKI.

2. Materials and methods 2.1. Data source The current study relies on the American College of Surgeons NSQIP database. The NSQIP was specifically developed to assess the quality of surgical care, and collects perioperative data on 135 variables, including preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity for patients undergoing major surgical procedures in both the inpatient and outpatient setting across the United States. Trained surgical clinical

reviewers prospectively collect the NSQIP data at each involved institution. Validated data from medical charts allows quantification of 30-day, risk-adjusted surgical outcomes, including postdischarge information. In 2011, the NSQIP included data from 315 participating sites and more than 1.7 million of cases had been contributed by 2011 [15,18]. 2.2. Study population Patients undergoing laparoscopic or open RN (Current Procedural Terminology codes: 50220, 50225, 50230, 50545, 50546) or PN (50240, 50542, 50543) between 2005 and 2011 with a postoperative diagnosis of malignant neoplasm of the kidney, except pelvis (International Classification of Diseases 9th edition code: 189.0), were identified within the NSQIP data set. Preoperative estimated glomerular filtration rate (GFR) was calculated using the Modification of Diet in Renal Disease formula, a 4-variable equation consisting of age, sex, ethnicity, and SCr levels [19,20]. A total of 548 cases (13.5%) were excluded because of missing data elements needed to estimate preoperative kidney function (age [n = 16, 0.4%], race [n = 397, 9.8%], sex [n = 8, 0.2%], and preoperative SCr levels [n = 171, 4.2%]), as previously reported [21]. Patients with a preoperative diagnosis of ARF—defined in the NSQIP as a rising SCr 4 3 mg/dl and increasing azotemia (increase in blood urea nitrogen [BUN]) in the 24 hours before surgery (n ¼ 15, 0.4%), requirement of dialysis in the 2 weeks before surgery (n ¼ 98, 2.8%) or an estimated GFR o30 ml/min/1.73 m2 (n ¼ 50, 1.5%)—were excluded from the analyses. This resulted in a final population of 3,320 patients. 2.3. Covariates For each patient, age at surgery, gender, body mass index (BMI), race, smoking status, alcohol consumption, preoperative SCr levels, BUN, hematocrit, and type of kidney surgery were available. Preoperative American Society of Anesthesiologists (ASA) score and functional health status, as well as history of hypertension (HTN), diabetes mellitus (DM), pulmonary (chronic obstructive pulmonary disease, ventilation dependency, and pneumonia), cardiac (congestive heart failure, myocardial infarction, percutaneous coronary intervention, and cardiac surgery), liver (esophageal varices and ascites), neurovascular (transient ischemic attack and stroke with/without neurological deficit), and peripheral vascular disease, was reported. Multiple imputation was used to estimate missing data for the following covariates: BMI (n ¼ 24), preoperative hematocrit (n ¼ 36), and preoperative BUN (n ¼ 116). 2.4. Outcomes The primary end point was the occurrence of 30-day AKI, defined as a rise in SCr level of 42 mg/dl from the

M. Schmid et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266

preoperative value without requirement for dialysis, or postoperative worsening of renal function requiring any form of renal replacement therapy—according to the NSQIP definitions for progressive renal insufficiency or ARF (NSQIP does not provide postoperative SCr values). Additional outcomes consisted of prolonged operative time (pOT), prolonged length of stay (pLOS), occurrence of complications classified according to previously reported methodology [22], blood transfusion and mortality. pOT and pLOS were defined as a length of operative time and hospital stay beyond the 75th percentile (Z206 min and Z5 d, respectively). Overall complications were defined as occurrence of any reported complication (including cardiovascular, pulmonary, septic, thromboembolic, wound complications, or urinary tract infection) according to previously reported methodology [22]. 2.5. Statistical analyses Descriptive statistics of categorical variables focused on frequencies and proportions. Means and standard deviations, medians, and interquartile ranges were reported for continuously coded variables. Multivariable logistic regression was used to examine the association between preoperative covariates and the risk of 30-day AKI. Covariates consisted of age at surgery, race, gender, BMI, smoking status, baseline comorbidities (including pulmonary disease, cardiac disease, HTN, DM, neurovascular disease, and systemic steroid treatment), preoperative CKD stage, and type and approach of surgery. All statistical analyses were performed using the R statistical package (R Foundation for Statistical Computing, Vienna, Austria), with a 2-sided significance level set at P o 0.05. An institutional review board waiver was obtained before conducting this study, in accordance with institutional regulation when dealing with deidentified administrative data.

1261

and to have an ASA score Z3, compared with those who underwent PN (all P o 0.001). Patients with preoperative normal renal function or CKD stage I were more commonly represented in the PN group than in the RN group (38.7 vs. 29.5%, P o 0.001), as opposed to patients with CKD stage II or III, who were more commonly represented in the RN group (70.5 vs. 61.3%, P o 0.001). Overall, 28 (0.8%) patients died within 30 days following surgery (Table 2). 3.2. Short-term outcome of renal function Overall, 60 patients (1.8%) experienced AKI with a mean time to presentation of 5.4 days (range: 1–25) following surgery. Relative to patients with normal postoperative kidney function, patients presenting AKI were more likely to have undergone RN than PN (73.3% vs. 26.7%, P ¼ 0.02) and more likely to have had open surgery than minimally invasive surgery (71.7% vs. 46.5%, P o 0.001). Mean preoperative SCr levels, estimated GFR, and BUN significantly differed between the patients who had AKI vs. those who did not (1.0 vs. 1.4 mg/dl, 80.9 vs. 60.1 ml/min/1.73 m2, and 16.9 vs. 22.0 mg/dl, respectively). More than half of the patients who suffered AKI had preoperative CKD stage III (56.7%) compared with only 20.6% of those without AKI (P o 0.001). 73.4% of patients with AKI had a preoperative ASA score ≥ 3 compared with 60.2% of patients without AKI (P o 0.03). Patients with AKI more frequently reported having required assistance from others for activities of daily living 30 days before surgery compared with those who did not experience AKI (5.0 vs. 0.4%, P o 0.001). Patients with AKI were more likely to have DM (37.8% vs. 19.8%, P ¼ 0.01), HTN (81.1% vs. 65.0%, P ¼ 0.04), neurovascular disease (24.3% vs. 4.9%, P o 0.001), or systemic steroid treatment (10.8% vs. 3.0%, P ¼ 0.01) (Table 1). Compared with no AKI, postoperative AKI was significantly associated with pLOS (Z5 d) (31.8% vs. 81.7%, P o 0.001), occurrence of any complication (6.5% vs. 43.3%, P o 0.001), and mortality (0.7% vs. 6.7%, P o 0.001) (Supplementary Table).

3. Results 3.3. Multivariable analysis 3.1. Baseline characteristics Overall, 1,944 (58.6%) and 1,376 (41.4%) patients treated with RN and PN between 2006 and 2011 were included, respectively. Overall, 1,559 (47.0%) and 1,761 (53.0%) underwent open and minimally invasive surgery, respectively. Table 1 shows the baseline characteristics of the study population. Mean and median age at surgery was 61.4 and 62 years, respectively. Most patients were men (61.3%) and white (88.8%). Overall, 1,107 (33.3%) patients were classified as having no preexisting renal impairment or CKD stage I, 1,509 (45.5%) had CKD stage II, and 704 (21.2%) had CKD stage III. Patients who underwent RN were more likely to be older than 75 years, to have a history of cardiac disease and HTN,

Table 3 shows the results of the multivariable logistic regression model testing the association between preoperative predictors and the occurrence of postoperative AKI. RN (odds ratio [OR] ¼ 2.87, P ¼ 0.02) and open surgery (OR ¼ 2.18, P ¼ 0.04) were independently associated with postoperative AKI. Obesity was associated with a 2.2-fold higher odd of 30-day loss of renal function (P ¼ 0.04). Other independent predictors included black race (OR ¼ 2.98, P ¼ 0.01), history of neurovascular disease (OR ¼ 5.29, P o 0.001), and systemic steroid treatment (OR ¼ 3.79, P ¼ 0.04). The strongest independent predictor of AKI was preoperative renal function: patients with CKD stage II and CKD stage III had a 10- and 26-fold higher odds of postoperative kidney function loss compared with those with

1262

M. Schmid et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266

Table 1 Characteristics of 3,320 patients with preoperative eGFR 430 ml/min/1.73 m2 at risk of acute kidney injury (AKI) after partial and radical tumor nephrectomy Variables

Total no. (%)

No AKI no. (%)

AKI no. (%)

No. of patients Age, y, median (IQR) 18–59 60–74 Z75

3,320 62 1,435 1,309 576

3,260 62 1,418 1,281 651

60 67.0 17 28 15

Procedure Partial nephrectomy Radical nephrectomy

1,376 (41.4) 1,944 (58.6)

1,360 (41.7) 1,900 (58.3)

16 (26.7) 44 (73.3)

0.019

Approach Open Laparoscopic

1,559 (47.0) 1,761 (53.0)

1,516 (46.5) 1,744 (53.5)

43 (71.7) 17 (28.3)

o0.001

Gender Male Female

2,094 (61.3) 1,226 (36.9)

2,050 (62.9) 1,210 (37.1)

44 (73.3) 16 (26.7)

0.097

Race White Black Other

1,948 (88.8) 284 (8.6) 88 (2.7)

2,905 (89.1) 270 (8.3) 85 (2.6)

43 (71.7) 14 (23.3) 3 (5.0)

o0.001

BMI, kg/m2, mean (SD) Underweight Normal Obesity (BMI 4 30)

30.5 32 1,727 1,557

(7.0) (1.0) (52.1) (47.0)

30.5 32 1,702 1,522

(7.0) (1.0) (52.3) (46.7)

32.4 0 25 35

(8.2) (0.01) (41.7) (58.3)

0.056 0.170

Functional health status Independent Partially dependent Totally dependent Unknown

834 16 1 2,469

(25.1) (0.5) (0.01) (74.7)

820 13 1 2,426

(25.2) (0.4) (0.01) (74.4)

14 3 0 43

(23.3) (5.0) (0.01) (71.7)

o0.001

ASA score 1 2 Z3

46 (1.4) 1,263 (38.0) 1,908 (60.5)

0 (0.01) 16 (26.7) 44 (73.4)

0.033

Baseline laboratory values Mean serum creatinine, mg/dl (SD) Mean eGFR, ml/min/1.73 m2 (SD) Mean BUN, mg/dl (SD) Mean HCT, % (SD)

1.0 80.5 16.9 40.0

(100) (53–70) (43.2) (39.4) (17.3)

(0.3) (27.1) (7.4) (5.2)

(98.2) (53–70) (43,5) (39.3) (17.2)

46 (1.4) 1,247 (38.3) 1,964 (60.2) 1.0 80.9 16.9 40.0

(0.3) (27.1) (7.3) (5.2)

1.4 60.1 22.0 40.1

(1.8) (58–73) (28.2) (46.7) (25.0)

P-value – 0.916 0.050

(0.4) (19.9) (8.0) (6.3)

o0.001 o0.001 o0.001 o0.001

CKD stage, ml/min/1.73 m2 None or stage I (eGFR 490) Stage II (eGFR 60–89) Stage III (eGFR 30–59)

1,107 (33.3) 1,509 (45.5) 704 (21.2)

1,103 (33.8) 1,487 (45.6) 670 (20.6)

4 (6.7) 22 (36.7) 34 (56.7)

o0.001

Comorbidities Pulmonary disease Liver disease Cardiac disease Hypertension (medication) Diabetes mellitus Neurovascular disease Peripheral vascular disease Systemic steroid treatment

122 4 248 1,491 458 120 23 72

118 4 243 1,651 444 111 23 68

4 0 5 30 14 9 0 4

(10.8) (0.1) (13.5) (81.1) (37.8) (24.3) (0.1) (10.8)

0.136 0.797 0.601 0.042 0.006 o0.001 0.536 0.007

2 (5.4) 5 (13.5)

0.644 0.247

Lifestyle Alcohol Smoking

(5.3) (0.2) (10.9) (65.3) (20.1) (5.3) (1.0) (3.2)

90 (3.9) 485 (21.2)

(5.3) (0.2) (10.8) (65.0) (19.8) (4.9) (1.0) (3.0)

88 (3.9) 480 (21.4)

Note: numbers in bold means they are significance (P o 0.05). eGFR ¼ estimated GFR (Modification of Diet in Renal Disease–formula); HCT ¼ hematocrit; IQR ¼ interquartile range; SD ¼ standard deviation.

M. Schmid et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266 Table 2 A total of 3,320 patients with preoperative estimated glomerular filtration rate 430 ml/min/1.73 m2 undergoing partial and radical tumor nephrectomy Variables

Partial nephrectomy no. (%)

Radical nephrectomy no. (%)

1,376 (41.4)

1,944 (58.6)

672 (48.8) 704 (51.2)

887 (45.6) 1,057 (54.4)

868 (63.1) 508 (36.9)

1,226 (63.1) 718 (36.9)

0.993

Age Z75, y Obesity, BMI Z 30 (kg/m2) CKD stage I II III

181 (13.2) 650 (47.5)

395 (20.3) 893 (46.3)

o0.001 0.476

533 (38.7) 593 (43.1) 250 (18.2)

574 (29.5) 916 (47.1) 454 (23.4)

o0.001

ASA score Z 3 Comorbidities Pulmonary disease Liver disease Cardiac disease Hypertension Diabetes mellitus Neurological disease Peripheral vascular disease

787 (57.2)

1,221 (62.8)

o0.001

77 2 80 855 248 45

122 6 173 1,301 388 85

No. of patients Approach Open Laparoscopic Gender Male Female

(8.2) (0.2) (8.8) (62.1) (18.0) (4.9)

8 (0.9)

P-value

Table 3 Multivariable logistic regression analysis for predictors of 30-day acute kidney injury following radical and partial nephrectomy in patients with a preoperative eGFR 430 ml/min/1.73 m2 Variables

– 0.068

(8.6) (0.4) (12.6) (66.9) (20.0) (6.2)

15 (1.1)

0.698 0.387 0.004 0.004 0.163 0.201 0.609

Note: numbers in bold means they are significance (P o 0.05).

normal preoperative renal function or CKD stage I (P ¼ 0.03 and 0.001, respectively).

1263

Odds ratio (OR)

95% CI

P-value

Age at surgery (ref. o75) Z75

1.35

0.61–2.99

0.456

Gender (ref. male) Female

0.56

0.26–1.23

0.147

Type of surgery (ref. partial nephrectomy) Radical nephrectomy

2.87

1.21–6.79

0.017

Approach (ref. laparoscopic) Open

2.18

1.04–4.57

0.039

Race (ref. white) Black Other

2.98 4.00

1.256–7.074 0.73–21.95

0.013 0.110

ASA (ref. o 3) Z3

1.12

0.47–2.66

0.803

1.3–77.27 3.44–204.24

0.027 0.002

0.036

CKD stage (ref. none or stage I) II III

10.001 26.49

Body weight, ref. BMI o 30 (kg/m2) Obesity (BMI Z 30)

2.24

1.06–4.76

Pulmonary disease Cardiac disease Hypertension Diabetes mellitus Neurovascular disease Systemic steroid treatment

1.89 0.64 0.86 1.92 5.29 3.79

0.6–5.95 0.23–1.82 0.34–2.17 0.91–4.07 2.21–12.65 1.1–13.09

0.280 0.405 0.755 0.087 o0.001 0.035

Note: numbers in bold means they are significance (P o 0.05). eGFR ¼ estimated GFR.

4. Discussion Cho et al. [23] showed in a cohort of Korean RN patients with preoperative normal kidney function that those who experienced AKI postsurgery had a 4.2-fold higher risk of new-onset CKD in the 1-year follow-up. Recent data have suggested that kidney function following renal surgery is correlated with overall survival, regardless of oncological outcomes [7,24,25]. Moreover, nephrotoxicity is a side effect of several treatment regimens for advanced RCC, including tyrosine kinase inhibitors. Taken together, the preservation of kidney function is an important end point in patients undergoing renal surgery. In the current study, we sought to examine the predictors of AKI following RN/PN in a large cohort of contemporary patients with RCC. The NSQIP provides detailed prospectively collected data on preoperative patients characteristics, including serum laboratory values, as well as 30-day postoperative outcomes and complications. Although the 30-day threshold for data collection of the NSQIP precluded us from examining long-term renal insufficiency, the incidence of

AKI is significantly associated with a decrease in short-term estimated GFR and the development of CKD at long-term follow-ups after surgery [14,26]. In this study, we showed that the risk of AKI following RN/PN was 1.8% of this NSQIP cohort. Previous studies have reported postoperative AKI or ARF rates or both ranging from 0.8% to 33.7% [23,27], varying according to institution, technique, approach, and method of data collection. Interestingly, we showed that 66.7% of patients undergoing RN/PN present with mildly or moderately reduced kidney function before surgery, which may expose them to an increased risk of 30-day acute renal loss of function, as demonstrated in the current study, but also to nonrenal complications, such as prolonged hospital stay, cardiovascular events, and mortality [28]. In comparison, using estimated GFR based on SCr levels, the incidence of CKD is estimated at 13% to 16% in the adult US population [29]. Additionally, up to 30% of patients undergoing PN or RN may have preexisting CKD, despite having normal preoperative SCr levels [5,9]. A recent study examining

1264

M. Schmid et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266

kidney dysfunction in general surgery patients showed that 64% of the cohort had reduced estimated GFR before surgery [21]. Given that preoperative renal function represents one of the strongest predictors of postoperative AKI, our results support the importance of assessing renal function before RN and PN. Third, our observations highlight the role of surgical approach on the risk of 30-day AKI. Indeed, RN was identified as an independent predictor for AKI compared with PN, where patients undergoing RN had approximately 3-fold increased odds of postoperative AKI. Better preservation of renal function has been hypothesized to improve overall survival after surgery for RCC and is a driving force behind the adoption of PN [24,25]. Furthermore, we observed that RN patients generally have reduced estimated GFR and higher CKD stages before surgery compared with that of their PN counterparts. These findings suggest that patients who may benefit the most from nephron-sparing surgery may not be deemed eligible for that approach. Nonetheless, it is possible that the poor outcomes recorded with RN are due to a selection bias, where younger and healthier patients with localized disease are more likely to receive nephron-sparing surgery. Additionally, the open approach was shown to have an adverse effect on postoperative renal function, independent of type of surgery (RN vs. PN). Particularly, patients receiving open surgery had 2.2-fold higher odds of AKI compared with their counterparts undergoing minimally invasive surgery. Despite reported equivalent long-term renal functional outcome [30], previous studies have suggested a potential negative effect of the minimally invasive approach on short-term renal function owing to prolonged warm ischemia time in comparison with the open approach [31]. Our findings, corroborated by Liu et al. [22], suggest that the minimally invasive approach is certainly not inferior to the open approach with regard to preservation of short-term renal function, though absence of data on tumor size precludes adjustment for case mix. Furthermore, the higher ratio of patients with elevated age and ASA score, comorbidities, and preexisting CKD III undergoing open surgery suggests that urologists may be preselecting for cases they perceive as more challenging; other patient characteristics that support this decision may be unknown and therefore cannot be adjusted for. Among medical comorbidities, in addition to CKD, obesity was identified as an independent predictor of AKI. Based on the fact that obesity has a negative long-term effect on kidney function [32,33], previous investigators had identified obesity as a risk factor for AKI after surgery [34]. Moreover, increasing BMI is associated with de novo CKD a year after PN [13]. Further predictors of AKI were neurovascular disease and “systemic steroid treatment,” defined as patients requiring regular administration of oral or parenteral corticosteroid medications for a chronic medical condition such as chronic obstructive pulmonary disease, rheumatoid arthritis, or inflammatory bowel disease—all

medical conditions that may be associated with renal impairment [35–38]. Interestingly, our study showed that HTN and DM are not independent predictors of AKI, despite their well-established association with long-term progressive renal disease [12,13]. Nonetheless, the presence of these comorbidities should guide the selection of surgery and approach to minimize the risk of renal injury. From a clinical perspective, our observations indicate the importance of identifying patients at risk for AKI before surgery and in the early postoperative setting. As AKI is independently associated with increased nonrenal complications, cardiovascular events, and consequently prolonged hospital stay and mortality [28], these adverse outcomes may be avoided or at least mitigated by careful patient preparation and postsurgery care. In addition, as AKI is a known risk factor for CKD [23], patients who experience AKI after renal surgery will require long-term follow-up of renal function. Patients themselves are often not aware of further risk factors for CKD; a previous report showed that, within a year after PN or RN, 16% of patients presented with new obesity and 31% with new-onset HTN, which was untreated in 13% of the cases at presentation [39]. Moreover, the incidence of AKI is not only undesirable for patients; it is also expensive, with direct health care costs of $8 billion annually [40]. Postoperative AKI increases the cost of hospitalization by an average of $28,359 [15]. Careful preoperative patient assessment is warranted (screening for CKD and risk factors) to avoid or minimize AKI in at-risk patients. Indeed, according to the NSQIP data, many patients undergo renal surgery without (recorded) data on preoperative renal function. We excluded 171 (4.2%) patients because of missing SCr data. Rates of missing SCr acquisition may range from 19% in NSQIP up to 81% in administrative databases [21,41]. Despite its strengths, our study was not devoid of limitations. Although several patient-specific factors were accounted for (age, sex, renal function before surgery, and medical comorbidities), many tumor (size, histology, grade, and stage) and surgical (warm or cold ischemia time and percentage of parenchyma resected) confounders are not accounted for in the NSQIP [7,42]. As such, it was not possible to use standardized preoperative stratification models such as the Motzer criteria or the RENAL score in this study [43]. Nonetheless, there were no significant differences concerning pOT and intraoperative blood transfusion rate on patients with AKI compared with those without suggesting equal tumor sizes and surgical complexity. Whereas intraoperative blood transfusion was significantly higher in RN compared with PN patients (14.4% vs. 6.0%, P o 0.001), the ratio of pOT was higher in PN patients (37.9% vs. 29.2%, P o 0.001). Furthermore performance status was not assessed by standardized tools for evaluation of patients with RCC, such as the Eastern Cooperative Oncology Group or Karnofsky score. Second, variables such as hospital and payer characteristics are not reported in the NSQIP, which precluded us from examining

M. Schmid et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266

the effect of hospital volume or other demographic factors such as insurance status. Third, data on postoperative SCr were not recorded. This parameter would have allowed better approximation of the severity of AKI; the NSQIP, referring to its definition of postoperative renal dysfunction by a SCr rise 42 mg/dl, has been criticized for potentially underestimating the incidence of mild and moderate AKI [44]. However, AKI has been evaluated in different surgical setting before in the NSQIP [45,46]. In general, there are multiple definitions of AKI, which defines an abrupt loss of kidney function and largely replaced the term “acute renal failure,” now reserved for “severe AKI.” The RIFLE (riskinjury-failure-loss-end stage renal disease) criteria [47] was the first consensus conform definition for AKI, followed by its the Acute Kidney Injury Network classification, which defines AKI as an absolute increase in the SCr concentration of Z0.3 mg/dl from baseline within 48 hours [48]. Most recently the Kidney Disease/Improving Global Outcomes group revised the definition of AKI while retaining Acute Kidney Injury Network staging criteria [49]. Apart from AKI definitions, SCr alone, known for its biological individuality, is not an adequate marker of renal function measurement [50]. Indeed, assessment of postoperative estimated GFR, which was used by several surgical series before [5,12,23], would have been a more appropriate end point to evaluate renal function outcome. 5. Conclusion Overall, 1.8% of patients undergoing surgery for RCC experience postoperative AKI. Surgical technique has a substantial effect on the risk of AKI, as patients treated with RN and with an open approach are at a higher risk of postoperative AKI. The strongest independent predictor of AKI is preexisting CKD stage II or III. Preoperative evaluation of kidney function and identification of predictors for postoperative AKI are important components of preoperative individual risk assessment and counseling in these patients. Appendix A. Supporting Information Supplementary material cited in this article is available online at http://dx.doi.org/10.1016/j.urolonc.2014.05.002.

References [1] Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin 2013;63:11–30:[Epub 2013/01/22]. [2] Motzer RJ, Agarwal N, Beard C, et al. Kidney cancer. J Natl Compr Canc Netw 2011;9:960–77:[Epub 2011/09/16]. [3] Klatte T, Shariat SF, Remzi M. Systematic review and meta-analysis of perioperative and oncologic outcomes of laparoscopic cryoablation versus laparoscopic partial nephrectomy for the treatment of small renal tumors. J Urol 2013;191(5):1209–17:[Epub 2013/11/16].

1265

[4] Psutka SP, Feldman AS, McDougal WS, McGovern FJ, Mueller P, Gervais DA. Long-term oncologic outcomes after radiofrequency ablation for T1 renal cell carcinoma. Eur Urol 2013;63:486–92:[Epub 2012/09/11]. [5] Huang WC, Levey AS, Serio AM, et al. Chronic kidney disease after nephrectomy in patients with renal cortical tumours: a retrospective cohort study. Lancet Oncol 2006;7:735–40:[Epub 2006/09/02]. [6] Barlow LJ, Korets R, Laudano M, Benson M, McKiernan J. Predicting renal functional outcomes after surgery for renal cortical tumours: a multifactorial analysis. BJU Int 2010;106:489–92:[Epub 2009/12/31]. [7] Thompson RH, Lane BR, Lohse CM, et al. Renal function after partial nephrectomy: effect of warm ischemia relative to quantity and quality of preserved kidney. Urology 2012;79:356–60:[Epub 2012/ 02/09]. [8] Weight CJ, Larson BT, Fergany AF, et al. Nephrectomy induced chronic renal insufficiency is associated with increased risk of cardiovascular death and death from any cause in patients with localized cT1b renal masses. J Urol 2010;183:1317–23:[Epub 2010/02/23]. [9] Lane BR, Poggio ED, Herts BR, Novick AC, Campbell SC. Renal function assessment in the era of chronic kidney disease: renewed emphasis on renal function centered patient care. J Urol 2009; 182:435–43:[discussion 43-4. Epub 2009/06/16]. [10] Bydash JR, Ishani A. Acute kidney injury and chronic kidney disease: a work in progress. Clin J Am Soc Nephrol 2011;6:2555–7:[Epub 2011/10/08]. [11] Rewa O, Bagshaw SM. Acute kidney injury-epidemiology, outcomes and economics. Nat Rev Nephrol 2014;10(4):193–207:[Epub 2014/01/22]. [12] Choi YS, Park YH, Kim YJ, Kang SH, Byun SS, Hong SH. Predictive factors for the development of chronic renal insufficiency after renal surgery: a multicenter study. Int Urol Nephrol 2013;46 (4):681–6:[Epub 2013/09/28]. [13] Kopp RP, Mehrazin R, Palazzi K, Bazzi WM, Patterson AL, Derweesh IH. Factors affecting renal function after open partial nephrectomy-a comparison of clampless and clamped warm ischemic technique. Urology 2012;80:865–70:[Epub 2012/09/07]. [14] Shin S, Han Y, Park H, et al. Risk factors for acute kidney injury after radical nephrectomy and inferior vena cava thrombectomy for renal cell carcinoma. J Vasc Surg 2013;58:1021–7:[Epub 2013/04/18]. [15] American College of Surgeons, National Surgical Quality Improvement Program. [16] Cima RR, Lackore KA, Nehring SA, et al. How best to measure surgical quality? Comparison of the Agency for Healthcare Research and Quality Patient Safety Indicators (AHRQ-PSI) and the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) postoperative adverse events at a single institution. Surgery 2011;150:943–9:[Epub 2011/08/31]. [17] Koch CG, Li L, Hixson E, Tang A, Phillips S, Henderson JM. What are the real rates of postoperative complications: elucidating inconsistencies between administrative and clinical data sources. J Am Coll Surg 2012;214:798–805:[Epub 2012/03/17]. [18] Henderson WG, Daley J. Design and statistical methodology of the National Surgical Quality Improvement Program: why is it what it is? Am J Surg 2009;198(Suppl. 5):S19–27:[Epub 2009/11/13]. [19] Levey AS, Greene T, Kusek J, GB. A simplified equation to predict glomerular filtration rate from serum creatinine [abstract]. J Am Soc Nephrol 2000:155. [20] National Kidney Foundation: K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39:S1. [21] Gaber AO, Moore LW, Aloia TA, et al. Cross-sectional and casecontrol analyses of the association of kidney function staging with adverse postoperative outcomes in general and vascular surgery. Ann Surg 2013;258:169–77:[Epub 2013/03/13]. [22] Liu JJ, Maxwell BG, Panousis P, Chung BI. Perioperative outcomes for laparoscopic and robotic compared with open prostatectomy using

1266

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32] [33] [34]

[35]

[36]

M. Schmid et al. / Urologic Oncology: Seminars and Original Investigations 32 (2014) 1259–1266 the National Surgical Quality Improvement Program (NSQIP) database. Urology 2013;82:579–83:[Epub 2013/07/24]. Cho A, Lee JE, Kwon GY, et al. Post-operative acute kidney injury in patients with renal cell carcinoma is a potent risk factor for new-onset chronic kidney disease after radical nephrectomy. Nephrol Dial Transplant 2011;26:3496–501:[Epub 2011/03/17]. Kim SP, Thompson RH, Boorjian SA, et al. Comparative effectiveness for survival and renal function of partial and radical nephrectomy for localized renal tumors: a systematic review and meta-analysis. J Urol 2012;188:51–7:[Epub 2012/05/18]. Weight CJ, Lieser G, Larson BT, et al. Partial nephrectomy is associated with improved overall survival compared to radical nephrectomy in patients with unanticipated benign renal tumours. Eur Urol 2010;58:293–8:[Epub 2010/06/16]. Takagi T, Kondo T, Iizuka J, et al. Predictors for postoperative renal function after open partial nephrectomy: including postoperative biomarkers. Int J Urol 2012;19:823–8:[Epub 2012/05/10]. Tanagho YS, Kaouk JH, Allaf ME, et al. Perioperative complications of robot-assisted partial nephrectomy: analysis of 886 patients at 5 United States centers. Urology 2013;81:573–9:[Epub 2013/03/05]. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004;351:1296–305:[Epub 2004/09/24]. Coresh J, Selvin E, Stevens LA, et al. Prevalence of chronic kidney disease in the United States. J Am Med Assoc 2007;298:2038–47: [Epub 2007/11/08]. Gill IS, Kavoussi LR, Lane BR, et al. Comparison of 1,800 laparoscopic and open partial nephrectomies for single renal tumors. J Urol 2007;178:41–6:[Epub 2007/06/19]. Lane BR, Novick AC, Babineau D, Fergany AF, Kaouk JH, Gill IS. Comparison of laparoscopic and open partial nephrectomy for tumor in a solitary kidney. J Urol 2008;179:847–51:[discussion 52. Epub 2008/01/29]. Wickman C, Kramer H. Obesity and kidney disease: potential mechanisms. Semin Nephrol 2013;33:14–22:[Epub 2013/02/05]. Kramer H, Luke A. Obesity and kidney disease: a big dilemma. Curr Opin Nephrol Hypertens 2007;16:237–41:[Epub 2007/04/11]. Kelz RR, Reinke CE, Zubizarreta JR, et al. Acute kidney injury, renal function, and the elderly obese surgical patient: a matched casecontrol study. Ann Surg 2013;258:359–63:[Epub 2013/05/17]. Gupta H, Ramanan B, Gupta PK, et al. Impact of COPD on postoperative outcomes: results from a national database. Chest 2013;143:1599–606:[Epub 2013/01/05]. Goules AV, Tatouli IP, Moutsopoulos HM, Tzioufas AG. Clinically significant renal involvement in primary Sjogren's syndrome: clinical presentation and outcome. Arthritis Rheum 2013;65:2945–53:[Epub 2013/10/30].

[37] Lewis B, Mukewar S, Lopez R, Brzezinski A, Hall P, Shen B. Frequency and risk factors of renal insufficiency in inflammatory bowel disease inpatients. Inflamm Bowel Dis 2013;19:1846–51:[Epub 2013/05/22]. [38] Tutuncu S, Ziegler AM, Scheitz JF, et al. Severe renal impairment is associated with symptomatic intracerebral hemorrhage after thrombolysis for ischemic stroke. Stroke 2013;44:3217–9:[Epub 2013/09/05]. [39] Schwaiger B, Schmid M, Tenschert W, et al. Risk factors of renal function deterioration after (partial) tumor nephrectomy. J Urol 2013;189:e259. [40] Uchino S, Kellum JA, Bellomo R, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. J Am Med Assoc 2005;294:813–8:[Epub 2005/08/18]. [41] Stevens LA, Fares G, Fleming J, et al. Low rates of testing and diagnostic codes usage in a commercial clinical laboratory: evidence for lack of physician awareness of chronic kidney disease. J Am Soc Nephrol 2005;16:2439–48:[Epub 2005/06/03]. [42] Lane BR, Babineau DC, Poggio ED, et al. Factors predicting renal functional outcome after partial nephrectomy. J Urol 2008;180:2363–8: [discussion 8-9. Epub 2008/10/22]. [43] Kutikov A, Uzzo RG. The R.E.N.A.L. nephrometry score: a comprehensive standardized system for quantitating renal tumor size, location and depth. J Urol 2009;182:844–53:[Epub 2009/07/21]. [44] Bihorac A, Brennan M, Ozrazgat-Baslanti T, et al. National surgical quality improvement program underestimates the risk associated with mild and moderate postoperative acute kidney injury. Crit Care Med 2013;41:2570–83:[Epub 2013/08/10]. [45] Glance LG, Wissler R, Mukamel DB, et al. Perioperative outcomes among patients with the modified metabolic syndrome who are undergoing noncardiac surgery. Anesthesiology 2010;113:859–72: [Epub 2010/09/03]. [46] Kheterpal S, Tremper KK, Heung M, et al. Development and validation of an acute kidney injury risk index for patients undergoing general surgery: results from a national data set. Anesthesiology 2009;110:505–15:[Epub 2009/02/13]. [47] Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8:R204–12:[Epub 2004/08/18]. [48] Mehta RL, Kellum JA, Shah SV, et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007;11:R31:[Epub 2007/03/03]. [49] Group KAW: KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl 2012;2:1–138. [50] Levey AS, Perrone RD, Madias NE. Serum creatinine and renal function. Annu Rev Med 1988;39:465–90:[Epub 1988/01/01].