European Urology

1. Introduction

Kidney tumors 4–7 cm in size have recently been managed with open partial nephrectomy (OPN), with 5-yr oncologic outcomes comparable to radical nephrectomy [1], [2], and [3]. Increasing numbers of patients referred to our center with clinical stage pT1b–T2 tumors are requesting and undergoing laparoscopic partial nephrectomy (LPN). This gradual and careful recent expansion of indications of LPN to include clinical stage pT1b–T2 tumors warrants evaluation of the impact of tumor size on perioperative and pathologic outcomes.

This study aims to examine the technical feasibility and perioperative safety and efficacy of LPN for enhancing renal tumors >4 cm in size. Operative data, intraoperative and postoperative complications, pathological outcomes, and renal functional outcomes were assessed over a 30-d perioperative period. In order to provide perspective on LPN outcomes in the study group (tumors >4 cm), outcomes of LPN in two control groups (group 1 with tumors <2 cm and group 2 with tumors 2–4 cm) were evaluated. To our knowledge, this is the initial series in the literature specifically focusing on LPN for clinical stage T1b–T2 tumors >4 cm.

2. Methods

Data were reviewed from an Institutional Review Board–approved, prospectively maintained database for 425 consecutive patients who underwent LPN by a single surgeon (ISG) between September 1999 and December 2005. Inclusion criteria included a single, organ-confined, contrast-enhancing renal mass. Indications for surgery for tumors clinically suspected on computed tomography (CT) scan to be angiomyolipoma (AML) included tumor size >4 cm; chronic flank pain (ostensibly from the AML) requiring opioid analgesics; or prior spontaneous perirenal hemorrhage, thereby increasing the risk of future hemorrhage. Exclusion criteria included renal vein involvement, lymphadenopathy, or extrarenal tumor extension. Patients were stratified for analysis into three groups based on radiologic tumor diameter: control group 1: <2 cm; control group 2: 2–4 cm; and study group 3: >4 cm.

Three-dimensional CT was utilized to evaluate the size, location, and depth of the renal mass and also to assess renal anatomy and vasculature in all patients. Central tumors were defined as those abutting the central sinus fat or pelvicalyceal system. Surgical technique for LPN was identical for all tumors in all groups whether suspected to be malignant or benign [4]. Intraoperative frozen-section analyses of the partial nephrectomy tumor specimen were typically performed to verify negative tumor margins.

Renal function was assessed using serum creatinine levels and estimated glomerular filtration rate (eGFR) calculated using the Modification of Diet in Renal Disease 2 (MDRD2) equation [5]. Serum creatinine measurements were obtained preoperatively and at 30 d postoperatively. Kidney dysfunction was graded using the National Kidney Foundation Dialysis Outcomes Quality Initiative Clinical Practice Guidelines (K/DOQI) [6]. This system, based on MDRD2 eGFR, classifies patients into five categories: stage 1: glomerular filtration rate (GFR) ≥90; stage 2: GFR 60–89; stage 3: GFR 30–59; stage 4: GFR 15–29; and stage 5: GFR < 15.

The National Cancer Institute Common Toxicity Criteria v2.0 were employed to grade complication severity: grade 1: expectant management and/or oral medication; grade 2: intravenous medication or chest tube insertion; grade 3: interventional radiology, endoscopic management, intubation, or reoperation; grade 4: major organ resection or chronic disability; and grade 5: death [7]. Intraoperative hemorrhage was defined as hemorrhage necessitating open conversion or transfusion. Postoperative hemorrhage was defined as acute blood loss necessitating transfusion, angioembolization, or reoperation. Urine leak was defined as biochemically confirmed urine drainage that persisted >7 d.

Normal data were compared using analysis of variance (ANOVA) tests and student t tests, and expressed as means with standard deviations. The significance of intergroup differences in the one-way ANOVA tests were evaluated by comparing the ratio of the between and within-groups mean squares to a Fisher F-distribution. Asymmetric data, categorical variables, and pairwise comparisons were compared using the chi-square or the Fisher exact test and expressed as medians with corresponding data ranges. In all analyses, the null hypothesis was rejected when p ≤ 0.05.

3. Results

Control groups 1 and 2 and study group 3 composed 21%, 65%, and 14% of the cohort, respectively (Table 1). Percent of patients with American Society of Anesthesiologists (ASA) score ≥3 was highest in study group 3: 41%, 53%, and 58% of patients in groups 1, 2, and 3, respectively (p = 0.05). This was attributed to higher incidence of medical comorbidities in group 3. Tumor sizes averaged 1.5 cm, 2.9 cm, and 6 cm in each group, respectively. Group 3 tumors were more complex as reflected by increases in both central location (p < 0.001) and intraparenchymal depth (p < 0.001). This resulted in increased rates of pelvicalyceal repair (p = 0.004) and heminephrectomy (>30% resection; p < 0.001), as well as increased mean warm ischemia time (WIT; 38 min; p = 0.007). Estimated operative blood loss, total operative time, and duration of hospital stay were equivalent among groups (Table 2).

Pathologic data are shown in Table 3. The incidence of carcinoma was highest in group 2 (80%; p = 0.002), while benign tumors were most prevalent in group 3 (47%; p = 0.01) primarily due to higher incidence of AML (20%). Increased tumor size did not correlate with higher incidence of stage pT3a–T3b disease. The rates of positive tumor margins were 0%, 0.5%, and 6.5% in groups 1, 2, and 3, respectively (p = 0.19).

Table 4 contains 30-d renal functional outcome data. Mean preoperative serum creatinine levels were 1.0 mg/dl, 1.0 mg/dl, and 1.1 mg/dl in groups 1, 2, and 3, respectively. In each group, mean postoperative serum creatinine levels were 1.2 mg/dl, 1.3 mg/dl, and 1.4 mg/dl, with mean increases of 0.15 mg/dl, 0.23 mg/dl, and 0.26 mg/dl, respectively (p = 0.12). Mean preoperative eGFRs for groups 1, 2, and 3, respectively, were 80 ml/min, 77 ml/min, and 77 ml/min, which decreased postoperatively by 12 ml/min, 15 ml/min, and 25 ml/min, respectively (p = 0.01). Thus, postoperative decline in kidney function in study group 3 was detected only by comparing eGFRs, not by comparing serum creatinine levels.

Preoperatively, stage 3–5 chronic kidney disease (CKD) was present in 31%, 35%, and 44% of patients in groups 1, 2, and 3, respectively (p = 0.15). Postoperatively 36%, 39%, and 47% of patients in each group, respectively, had a ≥1 CKD stage increase (p = 0.26). Patients with preoperative stage 3–5 CKD were at highest risk for postoperative CKD stage increase, which occurred in 3%, 17%, and 25% of patients in each group, respectively (p < 0.001). Overall 52%, 52%, and 63% of patients in each group, respectively, were classified as having stage 3–5 CKD (p = 0.5). Mean WIT in patients without postoperative CKD stage increase was similar at 31 min, 31 min, and 32 min in groups 1, 2, and 3, respectively (p = NS). Mean WIT in patients with postoperative CKD stage increase was 31 min, 35 min, and 44 min in groups 1, 2, and 3, respectively (p < 0.001). In patients with a solitary kidney, postoperative CKD stage increase occurred in 50%, 45%, and 43% in each group, respectively (p = 0.6).

Intraoperative complications occurred in 9%, 8%, and 7% of patients in groups 1, 2, and 3, respectively (p = 0.4; Table 5). Compared to patients without any intraoperative complication, those experiencing an intraoperative complication had an increased relative risk of postoperative urologic complications (risk ratio [RR], 4.8; p = 0.001) and nonurologic complications (RR, 3.8; p < 0.001). All conversions were to OPN, and this occurred at rates of 1%, 2%, and 2% in groups 1, 2, and 3, respectively. Reoperation for control of postoperative hemorrhage was necessary in two patients (0.4%) in group 2 only.

Overall postoperative complication rates were 11%, 24%, and 24% in groups 1, 2, and 3, respectively (p = 0.03, Table 6). Genitourinary complication rates were equivalent at 6%, 10%, and 7% in each group, respectively (p = 0.4). Nongenitourinary complications occurred in 6%, 17%, and 17% of patients in groups 1, 2, and 3, respectively, and these complications were of greater severity in group 2 (p = 0.001). Compared to control groups 1 and 2, overall complication rates and severity of complications were not increased significantly in study group 3.

Comparison of LPN for pathologically confirmed malignant tumors versus benign tumors is shown in Table 7. There were no significant differences in operative time, WIT, and estimated operative blood loss among the groups. In study group 3 the preoperative and postoperative eGFR was significantly lower in the cancer group. There was no difference in the postoperative percent decrease in eGFR between malignant tumors versus benign tumors in group 2 and group 3 patients. Postoperative complication rates for cancer patients in each group were 9%, 27%, and 32% in groups 1, 2, and 3, respectively (p < 0.001), while postoperative complication rates for benign tumors remained constant at 15%, 16%, and 16%, respectively. Importantly, 72% of the postoperative complications in patients with malignant tumors in groups 2 and 3 were nongenitourinary in origin, indicating that the increase in complication rate was not technical or procedural in origin.

4. Discussion

At its inception, OPN was reserved for tumors <4 cm. As experience grew it became apparent that OPN could be utilized effectively for 4–7 cm tumors [8], [9], [10], and [11]. Similarly, LPN has been routinely conducted for stage T1a tumors (<4 cm), however its safety and efficacy for clinical stage T1b–T2 tumors has not been established. Tumors 4–7 cm in size are generally more complex and central, thus utilizing LPN for their removal could theoretically be associated with increased risk of operative adverse events. To assess these concerns we evaluated technical feasibility and perioperative outcomes of LPN in patients with tumors >4 cm, and compared this data with outcomes for tumors ≤4 cm. Consecutive patients with an enhancing renal mass constituted the study population. This report focuses specifically on perioperative outcomes within 30 d of surgery, presenting data on short-term renal functional outcomes and complication incidence and severity. Intermediate-term outcomes, 5-yr oncologic outcomes, and renal functional outcomes will be reported in an accompanying analysis for this cohort.

More study group 3 patients had an ASA score ≥3 (p = 0.05), deeper tumors (p < 0.001), and central tumors (p < 0.001). Tumors >4 cm more often required transperitoneal approach selection (p = 0.04), pelvicalyceal repair (p = 0.004), and heminephrectomy (p < 0.001). This was accomplished without significant increases in intraoperative complications, estimated blood loss, positive tumor margins, conversion rates, or duration of hospital stay. There was an increase in positive tumor margins in group 3 in 2 of 31 patients with malignancies and was not statistically significant. Analysis of the margin status is ongoing as our experience with LPN for stage pT1b tumors increases. A total of eight patients (2%) underwent conversion to laparoscopic radical nephrectomy, thus intraoperative issues were effectively managed laparoscopically in the vast majority of cases. Mean WIT varied by only 8 min among groups (range: 30–38 min) despite increased tumor complexity. Prolonged WIT did not correlate with a higher incidence of hemorrhage or urine leakage, and mean WIT in patients with and without postoperative complications was similar at 32 min.

Intraoperative and postoperative technical complication rates did not correlate with increases in tumor size or WIT. Our complication rates are commensurate with previously published OPN and LPN data [12], [13], [14], and [15]. Occurrence of an intraoperative complication increased the chance of a postoperative complication. This equated to relative risks that increased 4.2-fold for total complications, and 4.8-fold for genitourinary complications, and 3.8-fold for nongenitourinary complications. Application of a biological hemostatic agent in the resection bed during renorrhaphy [16] was introduced halfway through the study period. This modification significantly correlated with decreased genitourinary complication rates, and, in concert with our increasing experience, may account for improved outcomes, despite an increase in tumor size and complexity over time.

Preoperative stage 3–5 CKD was present in 31%, 35%, and 44% of patients in groups 1, 2, and 3, respectively (p = 0.15). A recent report highlighted that 27% of patients with a small renal mass had baseline stage 3 CKD at the time of diagnosis [17]. Similarly, a recent study from our institution found that 21% of patients presenting with a unilateral kidney tumor <4 cm in diameter, bilateral kidneys appearing normal on CT, and a normal serum creatinine levels actually had stage 3–5 CKD at baseline (unpublished data). As such, the use of eGFR instead of serum creatinine measurement is clinically more relevant, and, going forward, should arguably be adopted as the standard of care in any preoperative assessment for patients undergoing partial nephrectomy, open or laparoscopic.

Prolonged WIT is associated with increased incidence of postoperative kidney dysfunction [18], [19], and [20]. Patients with preoperative CKD have lower tolerance for prolonged WIT [19], and [21]. The analysis identified a subset of patients at higher risk for postoperative CKD stage progression. Those with tumors >4 cm and stage 3–5 CKD at baseline had an 8-fold increased relative risk of CKD stage progression (p < 0.001). Although our overall mean WIT of 32 min is within with the widely accepted 30-min limit, it is important to emphasize that the cohort reported herein reflects our initial LPN data and not our current LPN data. Our recently reported “early unclamping” LPN technique has decreased our WIT by >50% [22]. In our most recent 100 consecutive LPN patients, mean WIT has been <15 min, >90% of patients had WIT ≤20 min, and no patient reached a WIT of 30 min (unpublished data). These WIT data compare favorably with the existing OPN literature [23], [24], and [25]. We believe such reduction of WIT will further minimize any sequelae of warm ischemia in patients undergoing LPN, specifically those with preexisting CKD of stage 3–V.

Indications for surgery in patients with AML included widely accepted criteria such as tumor with diameter >4 cm, chronic severe flank pain, and/or prior spontaneous hemorrhage [17]. AMLs composed 20% of tumors in study group 3 compared to 10% and 5% in control groups 1 and 2, respectively (p = 0.003). In patients with pathologically confirmed AML, preoperative CT imaging showed presence of visible adipose tissue in the contrast-enhancing tumor in 10%, 29%, and 83% of tumors in groups 1, 2, and 3, respectively. Six tertiary referral patients with radiologically diagnosed AML and extreme patient anxiety for cancer specifically requested LPN despite being advised about surveillance, angioinfarction, and risks of surgery for their likely benign condition. One group 2 patient with AML underwent LPN to rule out cancer prior to contralateral kidney donation per institutional protocol. All group 3 patients with preoperatively CT-suspected AML (n = 12) underwent an indicated elective LPN to prevent recurrent bleeding episodes in the future. Regardless of tumor type the technique of LPN employed for tumor excision and renorrhaphy was identical in all tumors in all groups [4]. This is reflected by similar perioperative outcomes in AML patients compared to those with pathologically confirmed malignant tumors in the total cohort: operative time (3.7 h vs 3.5 h; p = 0.4), estimated blood loss (250 ml vs 237 ml; p = 0.8), WIT (32.2 min vs 32 min; p = 0.9), rate of positive tumor margins (0% vs 1%; p = 0.6), and postoperative complication rate (12.5% vs 18%; p = 0.5).

Limitations of this study include the retrospective nature of the data analysis, the small number of patients with pathologically confirmed renal cancer in group 3, and the acquisition of operative data from a single institution by a single surgeon. Our prospective database includes only patients who actually underwent LPN; thus, demographic and radiologic data on patients excluded from LPN and the reason for their exclusion are not available. However, in the senior author's tertiary practice, exclusion from LPN for technical reasons is rare at this writing: for example, in 2007, >200 LPNs were performed, and only three patients were deemed unfit for LPN and referred for OPN. It is recognized that significant laparoscopic experience is necessary before embarking on the advanced LPN procedures reported herein. As such, these data may not be directly transferable to the practicing urologist.

Our data indicate negligible differences in perioperative outcomes of LPN for tumors >4 cm compared to outcomes of LPN for tumors <4 cm. The fundamental concept of nephron preservation supersedes the technique with which it is accomplished, open or laparoscopic. If an indicated LPN cannot be performed or completed laparoscopically, the default procedure should preferably be an OPN rather than a laparoscopic radical nephrectomy. As experience with larger tumors increases at various institutions, LPN data will become more comprehensive and have wider applicability. There is a higher incidence of postoperative renal dysfunction in patients with tumors >4 cm, the causes of which are multifactorial and due in part to prolonged WIT. This study highlights the need to minimize WIT to times commensurate with OPN, a goal that has already been reached [22].

5. Conclusions

LPN for tumors >4 cm has similar operative efficacy and perioperative and pathologic success rate as LPN for smaller tumors. This is achieved without increasing the incidence of open conversion, reoperation, intraoperative or postoperative complications, or positive tumor margins. Surgeon experience and patient selection are paramount for success.