- Email: [email protected]
Functional and Oncological Outcomes of Partial Nephrectomy of Solitary Kidneys
Functional and Oncological Outcomes of Partial Nephrectomy of Solitary Kidneys Jeffrey La Rochelle, Brian Shuch, Stephen Riggs, Li-Jung Liang, Ardavan Saadat, Fairooz Kabbinavar, Allan Pantuck and Arie Belldegrun* From the Department of Urology (JL, BS, SR, AS, FK, AP, AB), School of Public Health (LJL) and Division of Hematology-Oncology, Department of Medicine (FK), David Geffen School of Medicine at University of California-Los Angeles, Los Angeles, California
Purpose: We examined outcomes after partial nephrectomy in patients with tumors in a solitary kidney to determine the extent to which patient, surgery and tumor specific variables influenced the glomerular filtration rate and local recurrence postoperatively. Materials and Methods: Demographics, renal function, comorbidities, renal cell carcinoma history, and operative and pathological data were recorded. The effect on changes in early and late postoperative glomerular filtration rate and local recurrence were analyzed. Results: In 84 patients undergoing a total of 89 partial nephrectomies the mean immediate postoperative decrease in the glomerular filtration rate in those with no ischemia, warm ischemia (mean 12 minutes) and cold ischemia (mean 33 minutes) was 29%, 37% and 45%, respectively (p ⬍0.01). Late glomerular filtration rate decreases were 12%, 6% and 16%, respectively (p ⫽ 0.17). Cold ischemia and multiple vascular risk factors were associated with immediate glomerular filtration rate decreases (p ⫽ 0.008 and 0.04, respectively). Local recurrence, which developed in 13 patients (18%), was associated with positive margins and T stage (p ⫽ 0.01 and 0.02, respectively). End stage renal disease developed in 3 patients (4%) and an additional 5 (6%) required nephrectomy for local recurrence. Conclusions: Partial nephrectomy generally results in a small decrease in the glomerular filtration rate, and limited warm and cold ischemia does not appear to adversely affect long-term renal function. Positive margins and T stage greater than 2 are the most important predictors of local recurrence in a solitary kidney. They pose a significant risk to end stage renal disease-free survival due to the need for completion nephrectomy in many of these patients. Partial nephrectomy should be considered the standard of care in all patients with tumor in the solitary kidney.
Abbreviations and Acronyms CKD ⫽ chronic kidney disease EBL ⫽ estimated blood loss ESRD ⫽ end stage renal disease GFR ⫽ glomerular filtration rate LR ⫽ local recurrence PN ⫽ partial nephrectomy RCC ⫽ renal cell carcinoma RN ⫽ radical nephrectomy VHL ⫽ von Hippel-Lindau disease Submitted for publication September 6, 2008. Study received institutional review board approval. * Correspondence: Division of Urologic Oncology, Department of Urology, Box 951738, CHS 66-134, Los Angeles, California 90095-1738 (telephone: 310-794-6584; e-mail: abelldegrun@mednet. ucla.edu)
Key Words: kidney; nephrectomy; kidney failure, chronic; carcinoma, renal cell; neoplasm recurrence, local AS more incidental small renal masses are discovered, more patients with a solitary kidney after initial nephrectomy are surviving for extended periods. The risk of a metachronous contralateral renal tumor does not significantly decrease with time1 and, thus, it is likely that the number of
patients with a tumor in a solitary kidney will increase. These patients are likely to have chronic kidney disease2 and a decreased GFR is associated with increased mortality and hospitalization.3 Therefore, it is increasingly important to understand the factors influencing the functional
0022-5347/09/1815-2037/0 THE JOURNAL OF UROLOGY® Copyright © 2009 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 181, 2037-2043, May 2009 Printed in U.S.A. DOI:10.1016/j.juro.2009.01.024
www.jurology.com
2037
2038
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
and oncological outcomes of PN. Patients with a solitary kidney serve as the ideal context in which to study these parameters. Because the patient is left with fewer nephrons, the risk of hyperfiltration injury is heightened and data are needed on renal function beyond the early postoperative period. Previous reports of the effects of PN used serum creatinine to measure renal function.4 – 6 However, creatinine is an imperfect metric for renal function and identical creatinine values in the same patient with time does not indicate equivalent renal function. The 4-variable Modification of Diet in Renal Disease equation that calculates GFR using serum creatinine, patient age, gender and race has been shown to be more accurate than serum creatinine alone.7 An easily accessible GFR calculator can be found online (http://nephron.org). Most patients with a solitary kidney have undergone previous RN for RCC, so that the LR rate is also of particular concern. If patients undergoing PN for a second primary RCC are likely to show tumors in the renal remnant due to a propensity for multifocal tumor formation, nephron sparing surgery may not be justified. We analyzed the outcomes of PN in solitary kidneys to determine which preoperative and postoperative variables were predictive of decreases in GFR and LR.
PATIENTS AND METHODS Using the institutional review board approved prospective University of California–Los Angeles kidney cancer database patients who underwent PN in an anatomically solitary kidney were identified. Patients undergoing ablative procedures were excluded from study. Recorded variables included patient age, gender, race, diabetes, hypertension, cardiovascular disease, time from prior RN, reason for prior RN, preoperative and postoperative GFR, use of arterial clamping, warm/cold ischemia time, EBL, collecting system entry, margin status, transfusion rate, urine leak, largest diameter (of the index tumor if multiple lesions), and tumor multifocality and position. A positive margin was defined as tumor cells at the inked surface of the tumor specimen on final pathological assessment. Patients with a positive margin did not undergo immediate completion nephrectomy but instead were observed with cross-sectional imaging. Intraoperative frozen sections were not routinely used. Hemostasis was typically achieved by suturing, argon beam coagulation and/or hemostatic agents. When arterial clamping was anticipated, 25 gm mannitol were administered before clamping. Tumor position was determined by operative report and defined as peripheral— exophytic lesions where the renal sinus or collecting system was not entered during resection, deep—lesions with an exophytic component where the collecting system or renal sinus was entered during resection and central—lesions without an exophytic component that could not be directly visualized after removing Gerota’s fascia and that extensively involved the renal sinus.
Renal functional outcomes included early and late postoperative GFR, calculated using the Modification of Diet in Renal Disease equation, GFR in ml per minute per 1.73 m2 ⫽ 186.3 ⫻ serum creatinine⫺1.154 ⫻ age⫺0.203 ⫻ 1.212 (if the patient is black) ⫻ 0.742 (if the patient is female). Nadir GFR was the lowest GFR in the immediate postoperative period and late GFR was any value beyond 1 month after surgery. These values were compared to preoperative GFR to determine the GFR decrease, which is expressed as a percent of the preoperative value. The development of CKD was not a primary end point because most patients met the criteria for CKD (GFR less than 60 ml per minute per 1.73 m2) before undergoing PN. ESRD was defined as permanent dialysis dependence. GFR in patients with ESRD was designated as zero and the percent decrease in GFR in these patients was 100%. The primary oncological outcome of interest in patients with RCC was LR-free survival with LR defined as any new tumor developing in the renal remnant after PN. A large percent of patients undergoing PN had prior large contralateral tumors or a history of metastatic disease, so that disease specific survival was only a secondary outcome. The rate of urine leak was determined and correlated with tumor and operative variables. Descriptive statistics are presented as the mean ⫾ SD, median and range or percent. Continuous variables are summarized as the mean ⫾ SD. They were compared using Student’s t and the Wilcoxon rank sum tests, as appropriate. Categorical variables were compared using the chi-square and Fisher exact tests, as appropriate. Survival curves were generated by the Kaplan-Meier method and group differences were assessed by the log rank test. Multiple regression models were used to examine whether any variables were significantly associated with 1) an immediate GFR decrease, 2) a late GFR decrease in patients with long-term followup data (between 1 month and 5 years) and 3) LR. Patients with VHL lesions or transitional cell carcinoma were excluded from LR and renal function (ESRD and GFR) analysis because of expected disease progression. Patients with LR were excluded from renal function analysis to remove the effects of additional procedures from the analysis. A cubic spline regression model with random effects was used to evaluate the GFR trajectory with time in patients without LR or immediate ESRD. Models also included patient level random effects to account for the correlation between repeated measures at preoperative and several postoperative time points. Since the number of data points per patient was different and the data points were not equally or closely spaced in the study, cubic spline functions, which are piecewise cubic polynomials, were used to estimate highly curved regression functions. All statistical tests were 2-tailed with a significance level of 0.05. Statistical analysis was done using SAS® version 9.1.
RESULTS In 84 patients with a solitary kidney a total of 89 PNs were performed between 1991 and 2007. Median clinical followup was 40 months (range 0.2 to 148) and 26 patients (31%) were followed more than 5 years. Median followup for renal function data was
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
140 120 100
14.7 months (range 0.03 to 120). Median patient age at surgery was 62 years. Median time between radical nephrectomy and PN was 81 months (range 1 to 536). Median preoperative GFR was 54.4 ml per minute per 1.73 m2. A total of 80 patients had previously undergone RN for a mass (73) or for benign causes (7) and 4 had a congenital solitary kidney. One patient had VHL and 1 had transitional cell carcinoma. Of the 72 patients with RCC 21 (29%) had prior or concurrent metastatic disease. There were 2 to 16 tumors in a total of 25 cases (28%). Mean tumor size was 3.8 cm (median 3, range 0.8 to 18). The position of the index tumor, which was known in 87 of 89 cases, was peripheral in 35 (40%), deep in 38 (44%) and central in 14 (16%). Mean EBL was 739 cc (median 600, range 50 to 2,500). EBL correlated with tumor size (p ⫽ 0.001) but not with multifocality. A total of 36 patients (40%) underwent transfusion intraoperatively or postoperatively. Decreased EBL was not associated with arterial clamping because clamping was generally done after substantial blood loss, rather than at the beginning of resection due to our willingness to allow higher blood loss in return for shorter clamp time. Arterial clamping was done during 34 PNs (38%), of which 12 (35%) were cooled. Mean cold ischemia time was 33 minutes (range 20 to 55). Mean warm ischemia time was 12 minutes (range 3 to 26). Urine leaks developed in 15 cases (17%). Leaks correlated with tumor size (p ⫽ 0.04) and position (peripheral vs deep or central p ⫽ 0.03). Arterial clamping was not associated with fewer urine leaks. Table 1 lists the results of pathological evaluation. T stage and Fuhrman grade were known in 76
80
89 76 14 (18) 41 (54) 21 (28) 0 76 56 (74) 3 (4) 13 (17) 4 (5)
GFR
Total Grade: 1 2 3 4 T stage (%): 1 2 3a 3b
GFR Changes The mean immediate GFR decrease in patients without arterial clamping was 29% (range ⫺26% to 100%) compared to 37% (range 0% to 100%) in patients with warm ischemia and 45% (range 9% to 100%) in those with cold ischemia (p ⬍0.01). The mean late GFR decrease (later than 1 month) in the groups with no clamping, warm ischemia and cold ischemia was 12%, 6% and 16%, respectively (p ⫽ 0.17). Four of 7 patients who required dialysis immediately postoperatively underwent arterial clamping, including 2 with a warm ischemia time of 10 and 17 minutes, and 2 with a cold ischemia time of 45 and 55 minutes, respectively. One of the 2 patients who remained dialysis dependent did not undergo arterial clamping and 1 had experienced 10 minutes of warm ischemia time. Of 21 patients with a preoperative GFR of greater than 60 ml per minute per 1.73 m2 16 (76%) and 31 (70%) of 44 with a GFR of greater than 45 ml per minute per 1.73 m2 maintained a GFR above those respective cutoffs. Long-term GFR values were plotted (fig. 1). No trend suggesting a marked progressive decrease in renal function was seen during the observation period using the cubic spline function.
60
77 (87) 67 7 1 2 11 (12) 8 3 1 (1)
40
RCC: Clear cell Papillary Chromophobe Unclassified Benign: Oncocytoma Angiomyolipoma Transitional cell Ca
20
No. Pts (%)
of 77 RCC cases. The 13 positive margins (17%) were not associated with tumor size, although there was a borderline significant association with central or deep vs peripheral tumors (p ⫽ 0.07). No gross tumor was left behind in any case. There were no cases of false-negative frozen sections.
0
Table 1. Pathological features
2039
0
10
20
30
40
50
60
Time (Months)
Figure 1. Longitudinal GFR in patients without LR using cubic spline function. Dotted lines indicate 95% CI.
2040
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
Table 2. Multiple regression analysis of postoperative immediate and late GFR decrease % Immediate Decrease (p value)
Continuous variables: Preop GFR Age at surgery Tumor size Cold ischemia time (mins) Warm ischemia time (mins) Dichotomized variables: Diabetes mellitus (yes vs no) Cardiovascular disease (yes vs no) Multifocality Tumor position (deep or central vs peripheral) No. risk factors: 3 (vs 1 or less) (0.04) 2 (vs 1 or less)
% Last Decrease
Estimate
95% CI
Estimate
95% CI*
⫺0.11 ⫺0.03 ⫺0.56 0.98 (0.0081) 0.69
⫺0.65–0.43* ⫺0.65–0.58* ⫺3.04–1.92* 0.26–1.70 (⬍0.01) ⫺0.64–2.02*
⫺0.39 ⫺0.19 ⫺0.42 0.22 ⫺0.009
⫺1.01–0.24 ⫺0.90–0.52 ⫺3.29–2.45 ⫺0.62–1.05 ⫺1.55–1.53
⫺21.8–31.0* ⫺12.1–46.3* ⫺27.8–4.00* ⫺16.1–17.3*
4.60 17.1 ⫺11.9 0.61 61.0 9.7
2.57–119.4 (0.04) ⫺19.4–38.8*
13.0 24.9 ⫺17.6 0.21
⫺17.6–43.5 ⫺8.89–58.6 ⫺36.0–0.75 ⫺19.1–19.5
67.4 21.2
⫺0.15–135 ⫺12.4–54.9
* p not significant.
0.6
Positive Margins No Yes
0.4
Survival Distribution Function
0.8
1.0
End Stage Renal Disease Seven patients (8%) required dialysis in the immediate postoperative period. Only 2 patients (2%) remained dialysis dependent. ESRD developed in 8 patients (10%), which was due in 5 to completion nephrectomy for LR. No patient had ESRD as a
Log-Rank Test P-value = 0.0002
0.2
Local Recurrence Of 71 RCC cases without VHL a first LR, defined as new tumor in the renal remnant, was observed in 13 (18%) at a median of 16.3 months (range 1.6 to 118). Four patients underwent immediate completion nephrectomy, 5 underwent repeat PN, 1 underwent ablation and 3 received no intervention. All LR cases were clear cell RCC. Stratifying LR-free survival by margin status demonstrated a highly significant difference (fig. 2). However, tumor multifocality, prior nephrectomy for RCC and prior metastatic disease did not predict a higher LR rate. Observed LRs after a positive margin occurred at a mean of 19 months, while those after a negative margin occurred at a mean of 32 months. Estimated 5-year LR-free survival in patients with negative and positive margins was 83% and 20%, respectively (p ⬍0.001). Univariate and Cox multiple regression analysis performed to predict LR included the variables margin status, T stage (1 or 2 vs 3), prior RN for RCC, time from RN to PN, tumor multifocality and tumor size (table 3). Grade was not included because all recurrences
were in grade 2 lesions, although grade 3 lesions comprised 28% of RCC lesions in the cohort. Only margin status and T stage were predictive on multivariate analysis (HR 7.4, p ⫽ 0.01 and HR 5.7, p ⫽ 0.02, respectively). Interestingly the patient with 16 papillary RCC lesions had no recurrence at a followup of 75 months.
0.0
Regression analysis was performed to assess immediate postoperative and late GFR effects of patient age, preoperative GFR, largest tumor dimension, hypertension, diabetes and cardiovascular disease, multifocality, tumor position and arterial clamping (table 2). GFR changes were on a continuum and continuous variables were assessed for the estimated change that they caused. Cold ischemia time and the simultaneous presence of all 3 vascular risk factors (diabetes, hypertension and cardiovascular disease) were the only variables significantly associated with a decrease in GFR immediately postoperatively (p ⫽ 0.008 and 0.04, respectively). For late GFR changes no variable was significantly associated with a GFR decrease.
0
20
40
60
80
100
120
Months
Figure 2. Time to first LR stratified by margin status. Dotted lines indicate 95% CI.
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
Table 3. Cox multiple regression model of time to first LR Variable
HR
95% CI
p Value
Pos margin T stage (3 vs 1 or 2) Previous RCC Multifocality Tumor size
7.4 5.7 1.07 1.5 0.94
1.629–33.801 1.345–24.360 0.179–6.330 0.233–9.221 0.652–1.345
0.0096 0.0182 0.9440 0.6833 0.7219
result of repeat PN. Three of 69 patients (4.3%) without LR had ESRD, including 2 immediately and 1 who did not require immediate postoperative dialysis but showed ESRD at 80 months. Estimated 1, 2, 5 and 10-year ESRD-free survival rates in all patients were 94%, 92%, 83% and 66%, respectively. In patients without LR estimated 1, 2, 5 and 10-year ESRD-free survival rates were 97%, 97%, 97% and 73%, respectively (fig. 3). Patients without LR were stratified by preoperative GFR (table 4). Two of 3 patients with a GFR of less than 35 ml per minute per 1.73 m2 had ESRD immediately in 1 and at 80 months in 1. Of 43 patients with a preoperative GFR of 35 to 60 ml per minute per m2 ESRD occurred immediately in 1. None of the 23 patients with a GFR of greater than 60 ml per minute per 1.73 m2 had ESRD (p ⬍0.01).
0.6 0.4
Local Recurrence No Yes
0.2
Survival Distribution Function
0.8
1.0
Survival Estimated 5-year disease specific survival in patients with RCC without metastasis was 89% compared to 54% in those with a history of distant disease (p ⬍0.001). In the group with prior distant disease 5-year disease specific survival was esti-
0.0
Log-Rank Test: p-value = 0.0023
0
20
40
60
80
100
120
140
Months
Figure 3. ESRD-free survival stratified by LR status. Dotted lines indicate 95% CI.
2041
Table 4. ESRD-free survival in patients without LR by preoperative GFR Preop GFR (ml/min/1.73 m2) Less than 35 35–60 60 or Greater
No. Pts
No. ESRD (%)
3 43 23
2 (66.7) 1 (2.3), p ⬍0.01 0
mated to be 80% in those rendered disease-free after PN vs 35% in those with persistent distant metastatic sites but this difference did not attain significance due to small sample size (p ⫽ 0.15).
DISCUSSION In this analysis of PN performed for tumors in a solitary kidney several interesting results were found. 1) Most patients with tumors in a solitary kidney presented with evidence of chronic kidney disease. 2) Cold ischemia and multiple risk factors for vascular disease were significantly associated with the degree of the decrease in GFR immediately postoperatively but no variable was associated with the degree of decrease in late GFR values. 3) Positive margins were significantly associated with LR and the single most important risk factor for ESRD was LR. 4) No pronounced decrease in renal function after PN was seen during the observation period and most patients who underwent PN in a solitary kidney were able to avoid longterm dialysis, provided that they did not experience LR and subsequent nephrectomy. The results of multiple regression analysis for postoperative GFR changes seem counterintuitive, in that cold ischemia time was associated with immediate GFR decreases, while warm ischemia was not. However, average cold ischemia time was significantly longer than warm ischemia time (33 vs 12 minutes). No patient had a warm ischemia of longer than 25 minutes and no patient with cold ischemia had a clamp time of less than 20 minutes. This is similar to findings in a recent multi-institutional study, in which a greater risk of short-term acute renal failure was also seen in patients with cold vs warm ischemia.4 The limit for safe cold ischemia time in that study was 35 minutes, which was exceeded in half of the patients with cold ischemia in our study. Likewise, in the previous series the stated limit for safe warm time was 20 minutes, which was exceeded by only 2 of 22 study patients with warm ischemia in our study. Our results are consistent with the findings of that study that estimated safe warm and cold ischemia times at 20 and 35 minutes, respectively, although slightly longer warm ischemia time may also be safe. The fact that most patients presenting with tumor in a solitary kidney had baseline renal insuffi-
2042
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
ciency in this series is consistent with the findings of Huang et al, who found only a 30% 3-year probability of maintaining a GFR of greater than 60 ml per minute per 1.73 m2 after RN.2 Fortunately patients with a solitary kidney are able to avoid ESRD after PN and most patients without CKD (GFR greater than 60 ml per minute per 1.73 m2) remained so. Patients with the lowest GFR (less than 35 ml per minute per 1.73 m2) are at higher risk for ESRD but 2 of the 3 in this category who did not experience LR were able to remain dialysis-free for at least 80 months. Therefore, no GFR cutoff should be used to exclude candidates for PN, although patient counseling regarding possible ESRD is important in those with a low GFR. Renal function does not appear to markedly decrease in the intermediate term in patients who do not show ESRD immediately. The single patient with progression to ESRD at 80 months had diabetes with severe cardiovascular disease. He underwent repeat coronary bypass grafting between PN and the development of ESRD. Our results show that the greatest threat to ESRD-free survival is LR. Five of the 13 patients with LR went on to ESRD due to completion nephrectomy and margin status was the factor with the greatest effect on LR. Elective PN series have not shown that positive margins have this same significant effect.8,9 A possible reason for this difference is that lesions in this series were not ideal tumors for PN. Of RCC cases 17 (23%) involved T3 lesions, which is a higher rate than in many other series.5,8,9 Three of these patients had positive margins and recurrence. The series by Fergany et al with a comparable rate of T3 lesions did not describe the effect of positive margins on LR.6 Furthermore, 60% of the tumors in the current series were deep or central. It is likely that positive margins in this study were more significant due to more invasive lesions and a larger tumor-parenchyma interface. In fact, 10 of the 13 cases of positive margins were in
deep or central tumors. It is possible that positive margins in the deepest portion of the tumor bed were more likely to result in LR because this area is usually not fulgurated with an argon beam coagulator to avoid damaging sutures placed for hemostasis or collecting system repair. We did not determine the effect of LR on disease specific survival due to the many patients with previous metastatic disease or large contralateral tumors in our study population, which confounded analysis. This study was limited by a relatively small number of patients and particularly by the limited number of patients who underwent arterial clamping. It is possible that renal ischemia has a long-term effect on GFR that we could not detect at a statistically significant level. Our study has the advantage of being relatively enriched with higher T stage lesions, which allowed us to determine a difference in LR for positive margins in this higher risk population.
CONCLUSIONS PN for tumors in a solitary kidney results in small changes in long-term renal function and good local disease control even in patients with marginal preoperative renal function, multifocal tumors or a history of metastatic disease. However, an effort should be made to limit renal ischemia. Positive margins in solitary kidneys are more likely to be problematic than in elective cases, particularly in those with higher T stage lesions. Every effort should be made to achieve negative margins, including sending frozen sections or briefly using an arterial clamp to improve visualization, since the effects of positive margins appear to be more significant than the longterm effects of brief arterial occlusion. When final pathological results reveal a positive margin, immediate completion nephrectomy may not be justified but patients require close monitoring with a high degree of suspicion for LR.
REFERENCES 1. Rabbani F, Herr HW, Almahmed T and Russo P: Temporal change in the risk of metachronous contralateral renal cell carcinoma: influence of tumor and demographic factors. J Clin Oncol 2002; 20: 2370. 2. Huang WC, Levey AS, Serio AM, Snyder M, Vickers AJ, Raj GV et al: Chronic kidney disease after nephrectomy in patients with renal cortical tumors: a retrospective cohort study. Lancet Oncol 2006; 7: 735. 3. Go AS, Chertow GM, Fan D, McCulloch CE and Hsu C: Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004; 351: 1296.
4. Thompson RH, Frank I, Lohse CM, Saad IR, Fergany A, Zincke H et al: The impact of ischemia time during open nephron-sparing surgery on solitary kidneys: a multi-institutional study. J Urol 2007; 177: 471.
7. Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S et al: Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate. Ann Intern Med 2006; 145: 247.
5. Ghavamian R, Cheville JC, Lohse CM, Weaver AL, Zincke H and Blute ML: Renal cell carcinoma in the solitary kidney: an analysis of complications and outcome after nephron-sparing surgery. J Urol 2002; 166: 454.
8. Kwon EO, Carver BS, Snyder ME and Russo P: Impact of surgical margins in patients undergoing partial nephrectomy for renal cortical tumors. BJU Int 2006; 99: 286.
6. Fergany AF, Saad IR, Woo L and Novick AC: Open partial nephrectomy for tumor in a solitary kidney: experience with 400 cases. J Urol 2006; 175: 1630.
9. Yossepowitch O, Thompson RH, Leibovich BC, Eggener SE, Pettus JA, Kwon ED et al: Positive surgical margins at partial nephrectomy: predictors and oncological outcomes. J Urol 2008; 179: 2158.
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
2043
EDITORIAL COMMENT Changes in urological practice based on recent information can lessen the likelihood of facing a new tumor in a solitary kidney. PN is equivalent to RN for T1 renal cortical tumors less than 7 cm, of which 20% are benign and 25% are indolent. Most patients with renal tumors are older than healthy kidney donors (median age 61 years) and many have significant comorbidities, including CKD (26%), diabetes, hypertension, a high body mass index and cardiovascular disease. RN is an independent risk factor for CKD and decreased overall survival, and it carries an increased risk of a cardiovascular event after surgery. Despite stage migration to small, incidental renal masses (T1, median less than 4 cm) evidence from national databases indicates that PN remains
underused (less than 20% of T1 cases). The solitary kidney is susceptible to operative damage whether no ischemia, warm ischemia or cold ischemia is used. Early measurements of serum creatinine or the estimated GFR may underestimate long-term ischemic effects. Increased training in open or minimally invasive PN, increased active surveillance in the elderly population and in comorbidly ill patients with small tumors, and the restriction of RN to massive tumors is imperative. Paul Russo Urology Service Department of Surgery Memorial-Sloan Kettering Cancer Center New York, New York
REPLY BY AUTHORS While it is true that some functional loss is typically incurred by the kidney undergoing resection, our results add to the evidence that PN is well tolerated for solitary kidneys and should be considered the gold standard.1,2 Ablative therapies or observation, while certainly appropriate for the elderly or those with significant comorbidities, should not be pursued solely out of concern for functional loss of the solitary kidney. However, our results should introduce a note of caution regarding further expanding indications for PN. Preservation of renal function is clearly preferable3 but we must be careful as we proceed towards PN for larger and more difficult tumors. Some of these cases might be better managed with RN rather than PN if a negative surgical margin cannot be reliably achieved.
One might have the impression from recently published PN series that margin status is not particularly important. However, the majority of lesions previously treated with PN in these series have been small peripheral lesions that are likely to have been slow growing tumors in the short term. Positive margins in this population may not be as predictive of short-term future LR as they would be in larger or more centrally located tumors. For larger or more centrally located lesions PN should be performed with particular care, preferably by those experienced with resection of more difficult tumors, to achieve the same oncological and functional results seen with small peripheral tumors. Cases with a positive margin should be monitored closely with a high degree of suspicion for LR, which may then be salvaged with repeat PN, RN or ablative techniques as necessary.
REFERENCES 1. Ghavamian R, Cheville JC, Lohse CM, Weaver AL, Zincke H and Blute ML: Renal cell carcinoma in the solitary kidney: an analysis of complications and outcome after nephron-sparing surgery. J Urol 2002; 168: 454.
2. Adkins KL, Chang SS, Cookson MS and Smith JA: Partial nephrectomy safely preserves renal function in patients with a solitary kidney. J Urol 2003; 169: 79.
3. Huang WC, Elkin EB, Levey AS, Jang TL and Russo P: Partial nephrectomy versus radical nephrectomy in patients with small renal tumors—is there a difference in mortality and cardiovascular outcomes? J Urol 2009; 181: 55.
Purpose: We examined outcomes after partial nephrectomy in patients with tumors in a solitary kidney to determine the extent to which patient, surgery and tumor specific variables influenced the glomerular filtration rate and local recurrence postoperatively. Materials and Methods: Demographics, renal function, comorbidities, renal cell carcinoma history, and operative and pathological data were recorded. The effect on changes in early and late postoperative glomerular filtration rate and local recurrence were analyzed. Results: In 84 patients undergoing a total of 89 partial nephrectomies the mean immediate postoperative decrease in the glomerular filtration rate in those with no ischemia, warm ischemia (mean 12 minutes) and cold ischemia (mean 33 minutes) was 29%, 37% and 45%, respectively (p ⬍0.01). Late glomerular filtration rate decreases were 12%, 6% and 16%, respectively (p ⫽ 0.17). Cold ischemia and multiple vascular risk factors were associated with immediate glomerular filtration rate decreases (p ⫽ 0.008 and 0.04, respectively). Local recurrence, which developed in 13 patients (18%), was associated with positive margins and T stage (p ⫽ 0.01 and 0.02, respectively). End stage renal disease developed in 3 patients (4%) and an additional 5 (6%) required nephrectomy for local recurrence. Conclusions: Partial nephrectomy generally results in a small decrease in the glomerular filtration rate, and limited warm and cold ischemia does not appear to adversely affect long-term renal function. Positive margins and T stage greater than 2 are the most important predictors of local recurrence in a solitary kidney. They pose a significant risk to end stage renal disease-free survival due to the need for completion nephrectomy in many of these patients. Partial nephrectomy should be considered the standard of care in all patients with tumor in the solitary kidney.
Abbreviations and Acronyms CKD ⫽ chronic kidney disease EBL ⫽ estimated blood loss ESRD ⫽ end stage renal disease GFR ⫽ glomerular filtration rate LR ⫽ local recurrence PN ⫽ partial nephrectomy RCC ⫽ renal cell carcinoma RN ⫽ radical nephrectomy VHL ⫽ von Hippel-Lindau disease Submitted for publication September 6, 2008. Study received institutional review board approval. * Correspondence: Division of Urologic Oncology, Department of Urology, Box 951738, CHS 66-134, Los Angeles, California 90095-1738 (telephone: 310-794-6584; e-mail: abelldegrun@mednet. ucla.edu)
Key Words: kidney; nephrectomy; kidney failure, chronic; carcinoma, renal cell; neoplasm recurrence, local AS more incidental small renal masses are discovered, more patients with a solitary kidney after initial nephrectomy are surviving for extended periods. The risk of a metachronous contralateral renal tumor does not significantly decrease with time1 and, thus, it is likely that the number of
patients with a tumor in a solitary kidney will increase. These patients are likely to have chronic kidney disease2 and a decreased GFR is associated with increased mortality and hospitalization.3 Therefore, it is increasingly important to understand the factors influencing the functional
0022-5347/09/1815-2037/0 THE JOURNAL OF UROLOGY® Copyright © 2009 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 181, 2037-2043, May 2009 Printed in U.S.A. DOI:10.1016/j.juro.2009.01.024
www.jurology.com
2037
2038
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
and oncological outcomes of PN. Patients with a solitary kidney serve as the ideal context in which to study these parameters. Because the patient is left with fewer nephrons, the risk of hyperfiltration injury is heightened and data are needed on renal function beyond the early postoperative period. Previous reports of the effects of PN used serum creatinine to measure renal function.4 – 6 However, creatinine is an imperfect metric for renal function and identical creatinine values in the same patient with time does not indicate equivalent renal function. The 4-variable Modification of Diet in Renal Disease equation that calculates GFR using serum creatinine, patient age, gender and race has been shown to be more accurate than serum creatinine alone.7 An easily accessible GFR calculator can be found online (http://nephron.org). Most patients with a solitary kidney have undergone previous RN for RCC, so that the LR rate is also of particular concern. If patients undergoing PN for a second primary RCC are likely to show tumors in the renal remnant due to a propensity for multifocal tumor formation, nephron sparing surgery may not be justified. We analyzed the outcomes of PN in solitary kidneys to determine which preoperative and postoperative variables were predictive of decreases in GFR and LR.
PATIENTS AND METHODS Using the institutional review board approved prospective University of California–Los Angeles kidney cancer database patients who underwent PN in an anatomically solitary kidney were identified. Patients undergoing ablative procedures were excluded from study. Recorded variables included patient age, gender, race, diabetes, hypertension, cardiovascular disease, time from prior RN, reason for prior RN, preoperative and postoperative GFR, use of arterial clamping, warm/cold ischemia time, EBL, collecting system entry, margin status, transfusion rate, urine leak, largest diameter (of the index tumor if multiple lesions), and tumor multifocality and position. A positive margin was defined as tumor cells at the inked surface of the tumor specimen on final pathological assessment. Patients with a positive margin did not undergo immediate completion nephrectomy but instead were observed with cross-sectional imaging. Intraoperative frozen sections were not routinely used. Hemostasis was typically achieved by suturing, argon beam coagulation and/or hemostatic agents. When arterial clamping was anticipated, 25 gm mannitol were administered before clamping. Tumor position was determined by operative report and defined as peripheral— exophytic lesions where the renal sinus or collecting system was not entered during resection, deep—lesions with an exophytic component where the collecting system or renal sinus was entered during resection and central—lesions without an exophytic component that could not be directly visualized after removing Gerota’s fascia and that extensively involved the renal sinus.
Renal functional outcomes included early and late postoperative GFR, calculated using the Modification of Diet in Renal Disease equation, GFR in ml per minute per 1.73 m2 ⫽ 186.3 ⫻ serum creatinine⫺1.154 ⫻ age⫺0.203 ⫻ 1.212 (if the patient is black) ⫻ 0.742 (if the patient is female). Nadir GFR was the lowest GFR in the immediate postoperative period and late GFR was any value beyond 1 month after surgery. These values were compared to preoperative GFR to determine the GFR decrease, which is expressed as a percent of the preoperative value. The development of CKD was not a primary end point because most patients met the criteria for CKD (GFR less than 60 ml per minute per 1.73 m2) before undergoing PN. ESRD was defined as permanent dialysis dependence. GFR in patients with ESRD was designated as zero and the percent decrease in GFR in these patients was 100%. The primary oncological outcome of interest in patients with RCC was LR-free survival with LR defined as any new tumor developing in the renal remnant after PN. A large percent of patients undergoing PN had prior large contralateral tumors or a history of metastatic disease, so that disease specific survival was only a secondary outcome. The rate of urine leak was determined and correlated with tumor and operative variables. Descriptive statistics are presented as the mean ⫾ SD, median and range or percent. Continuous variables are summarized as the mean ⫾ SD. They were compared using Student’s t and the Wilcoxon rank sum tests, as appropriate. Categorical variables were compared using the chi-square and Fisher exact tests, as appropriate. Survival curves were generated by the Kaplan-Meier method and group differences were assessed by the log rank test. Multiple regression models were used to examine whether any variables were significantly associated with 1) an immediate GFR decrease, 2) a late GFR decrease in patients with long-term followup data (between 1 month and 5 years) and 3) LR. Patients with VHL lesions or transitional cell carcinoma were excluded from LR and renal function (ESRD and GFR) analysis because of expected disease progression. Patients with LR were excluded from renal function analysis to remove the effects of additional procedures from the analysis. A cubic spline regression model with random effects was used to evaluate the GFR trajectory with time in patients without LR or immediate ESRD. Models also included patient level random effects to account for the correlation between repeated measures at preoperative and several postoperative time points. Since the number of data points per patient was different and the data points were not equally or closely spaced in the study, cubic spline functions, which are piecewise cubic polynomials, were used to estimate highly curved regression functions. All statistical tests were 2-tailed with a significance level of 0.05. Statistical analysis was done using SAS® version 9.1.
RESULTS In 84 patients with a solitary kidney a total of 89 PNs were performed between 1991 and 2007. Median clinical followup was 40 months (range 0.2 to 148) and 26 patients (31%) were followed more than 5 years. Median followup for renal function data was
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
140 120 100
14.7 months (range 0.03 to 120). Median patient age at surgery was 62 years. Median time between radical nephrectomy and PN was 81 months (range 1 to 536). Median preoperative GFR was 54.4 ml per minute per 1.73 m2. A total of 80 patients had previously undergone RN for a mass (73) or for benign causes (7) and 4 had a congenital solitary kidney. One patient had VHL and 1 had transitional cell carcinoma. Of the 72 patients with RCC 21 (29%) had prior or concurrent metastatic disease. There were 2 to 16 tumors in a total of 25 cases (28%). Mean tumor size was 3.8 cm (median 3, range 0.8 to 18). The position of the index tumor, which was known in 87 of 89 cases, was peripheral in 35 (40%), deep in 38 (44%) and central in 14 (16%). Mean EBL was 739 cc (median 600, range 50 to 2,500). EBL correlated with tumor size (p ⫽ 0.001) but not with multifocality. A total of 36 patients (40%) underwent transfusion intraoperatively or postoperatively. Decreased EBL was not associated with arterial clamping because clamping was generally done after substantial blood loss, rather than at the beginning of resection due to our willingness to allow higher blood loss in return for shorter clamp time. Arterial clamping was done during 34 PNs (38%), of which 12 (35%) were cooled. Mean cold ischemia time was 33 minutes (range 20 to 55). Mean warm ischemia time was 12 minutes (range 3 to 26). Urine leaks developed in 15 cases (17%). Leaks correlated with tumor size (p ⫽ 0.04) and position (peripheral vs deep or central p ⫽ 0.03). Arterial clamping was not associated with fewer urine leaks. Table 1 lists the results of pathological evaluation. T stage and Fuhrman grade were known in 76
80
89 76 14 (18) 41 (54) 21 (28) 0 76 56 (74) 3 (4) 13 (17) 4 (5)
GFR
Total Grade: 1 2 3 4 T stage (%): 1 2 3a 3b
GFR Changes The mean immediate GFR decrease in patients without arterial clamping was 29% (range ⫺26% to 100%) compared to 37% (range 0% to 100%) in patients with warm ischemia and 45% (range 9% to 100%) in those with cold ischemia (p ⬍0.01). The mean late GFR decrease (later than 1 month) in the groups with no clamping, warm ischemia and cold ischemia was 12%, 6% and 16%, respectively (p ⫽ 0.17). Four of 7 patients who required dialysis immediately postoperatively underwent arterial clamping, including 2 with a warm ischemia time of 10 and 17 minutes, and 2 with a cold ischemia time of 45 and 55 minutes, respectively. One of the 2 patients who remained dialysis dependent did not undergo arterial clamping and 1 had experienced 10 minutes of warm ischemia time. Of 21 patients with a preoperative GFR of greater than 60 ml per minute per 1.73 m2 16 (76%) and 31 (70%) of 44 with a GFR of greater than 45 ml per minute per 1.73 m2 maintained a GFR above those respective cutoffs. Long-term GFR values were plotted (fig. 1). No trend suggesting a marked progressive decrease in renal function was seen during the observation period using the cubic spline function.
60
77 (87) 67 7 1 2 11 (12) 8 3 1 (1)
40
RCC: Clear cell Papillary Chromophobe Unclassified Benign: Oncocytoma Angiomyolipoma Transitional cell Ca
20
No. Pts (%)
of 77 RCC cases. The 13 positive margins (17%) were not associated with tumor size, although there was a borderline significant association with central or deep vs peripheral tumors (p ⫽ 0.07). No gross tumor was left behind in any case. There were no cases of false-negative frozen sections.
0
Table 1. Pathological features
2039
0
10
20
30
40
50
60
Time (Months)
Figure 1. Longitudinal GFR in patients without LR using cubic spline function. Dotted lines indicate 95% CI.
2040
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
Table 2. Multiple regression analysis of postoperative immediate and late GFR decrease % Immediate Decrease (p value)
Continuous variables: Preop GFR Age at surgery Tumor size Cold ischemia time (mins) Warm ischemia time (mins) Dichotomized variables: Diabetes mellitus (yes vs no) Cardiovascular disease (yes vs no) Multifocality Tumor position (deep or central vs peripheral) No. risk factors: 3 (vs 1 or less) (0.04) 2 (vs 1 or less)
% Last Decrease
Estimate
95% CI
Estimate
95% CI*
⫺0.11 ⫺0.03 ⫺0.56 0.98 (0.0081) 0.69
⫺0.65–0.43* ⫺0.65–0.58* ⫺3.04–1.92* 0.26–1.70 (⬍0.01) ⫺0.64–2.02*
⫺0.39 ⫺0.19 ⫺0.42 0.22 ⫺0.009
⫺1.01–0.24 ⫺0.90–0.52 ⫺3.29–2.45 ⫺0.62–1.05 ⫺1.55–1.53
⫺21.8–31.0* ⫺12.1–46.3* ⫺27.8–4.00* ⫺16.1–17.3*
4.60 17.1 ⫺11.9 0.61 61.0 9.7
2.57–119.4 (0.04) ⫺19.4–38.8*
13.0 24.9 ⫺17.6 0.21
⫺17.6–43.5 ⫺8.89–58.6 ⫺36.0–0.75 ⫺19.1–19.5
67.4 21.2
⫺0.15–135 ⫺12.4–54.9
* p not significant.
0.6
Positive Margins No Yes
0.4
Survival Distribution Function
0.8
1.0
End Stage Renal Disease Seven patients (8%) required dialysis in the immediate postoperative period. Only 2 patients (2%) remained dialysis dependent. ESRD developed in 8 patients (10%), which was due in 5 to completion nephrectomy for LR. No patient had ESRD as a
Log-Rank Test P-value = 0.0002
0.2
Local Recurrence Of 71 RCC cases without VHL a first LR, defined as new tumor in the renal remnant, was observed in 13 (18%) at a median of 16.3 months (range 1.6 to 118). Four patients underwent immediate completion nephrectomy, 5 underwent repeat PN, 1 underwent ablation and 3 received no intervention. All LR cases were clear cell RCC. Stratifying LR-free survival by margin status demonstrated a highly significant difference (fig. 2). However, tumor multifocality, prior nephrectomy for RCC and prior metastatic disease did not predict a higher LR rate. Observed LRs after a positive margin occurred at a mean of 19 months, while those after a negative margin occurred at a mean of 32 months. Estimated 5-year LR-free survival in patients with negative and positive margins was 83% and 20%, respectively (p ⬍0.001). Univariate and Cox multiple regression analysis performed to predict LR included the variables margin status, T stage (1 or 2 vs 3), prior RN for RCC, time from RN to PN, tumor multifocality and tumor size (table 3). Grade was not included because all recurrences
were in grade 2 lesions, although grade 3 lesions comprised 28% of RCC lesions in the cohort. Only margin status and T stage were predictive on multivariate analysis (HR 7.4, p ⫽ 0.01 and HR 5.7, p ⫽ 0.02, respectively). Interestingly the patient with 16 papillary RCC lesions had no recurrence at a followup of 75 months.
0.0
Regression analysis was performed to assess immediate postoperative and late GFR effects of patient age, preoperative GFR, largest tumor dimension, hypertension, diabetes and cardiovascular disease, multifocality, tumor position and arterial clamping (table 2). GFR changes were on a continuum and continuous variables were assessed for the estimated change that they caused. Cold ischemia time and the simultaneous presence of all 3 vascular risk factors (diabetes, hypertension and cardiovascular disease) were the only variables significantly associated with a decrease in GFR immediately postoperatively (p ⫽ 0.008 and 0.04, respectively). For late GFR changes no variable was significantly associated with a GFR decrease.
0
20
40
60
80
100
120
Months
Figure 2. Time to first LR stratified by margin status. Dotted lines indicate 95% CI.
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
Table 3. Cox multiple regression model of time to first LR Variable
HR
95% CI
p Value
Pos margin T stage (3 vs 1 or 2) Previous RCC Multifocality Tumor size
7.4 5.7 1.07 1.5 0.94
1.629–33.801 1.345–24.360 0.179–6.330 0.233–9.221 0.652–1.345
0.0096 0.0182 0.9440 0.6833 0.7219
result of repeat PN. Three of 69 patients (4.3%) without LR had ESRD, including 2 immediately and 1 who did not require immediate postoperative dialysis but showed ESRD at 80 months. Estimated 1, 2, 5 and 10-year ESRD-free survival rates in all patients were 94%, 92%, 83% and 66%, respectively. In patients without LR estimated 1, 2, 5 and 10-year ESRD-free survival rates were 97%, 97%, 97% and 73%, respectively (fig. 3). Patients without LR were stratified by preoperative GFR (table 4). Two of 3 patients with a GFR of less than 35 ml per minute per 1.73 m2 had ESRD immediately in 1 and at 80 months in 1. Of 43 patients with a preoperative GFR of 35 to 60 ml per minute per m2 ESRD occurred immediately in 1. None of the 23 patients with a GFR of greater than 60 ml per minute per 1.73 m2 had ESRD (p ⬍0.01).
0.6 0.4
Local Recurrence No Yes
0.2
Survival Distribution Function
0.8
1.0
Survival Estimated 5-year disease specific survival in patients with RCC without metastasis was 89% compared to 54% in those with a history of distant disease (p ⬍0.001). In the group with prior distant disease 5-year disease specific survival was esti-
0.0
Log-Rank Test: p-value = 0.0023
0
20
40
60
80
100
120
140
Months
Figure 3. ESRD-free survival stratified by LR status. Dotted lines indicate 95% CI.
2041
Table 4. ESRD-free survival in patients without LR by preoperative GFR Preop GFR (ml/min/1.73 m2) Less than 35 35–60 60 or Greater
No. Pts
No. ESRD (%)
3 43 23
2 (66.7) 1 (2.3), p ⬍0.01 0
mated to be 80% in those rendered disease-free after PN vs 35% in those with persistent distant metastatic sites but this difference did not attain significance due to small sample size (p ⫽ 0.15).
DISCUSSION In this analysis of PN performed for tumors in a solitary kidney several interesting results were found. 1) Most patients with tumors in a solitary kidney presented with evidence of chronic kidney disease. 2) Cold ischemia and multiple risk factors for vascular disease were significantly associated with the degree of the decrease in GFR immediately postoperatively but no variable was associated with the degree of decrease in late GFR values. 3) Positive margins were significantly associated with LR and the single most important risk factor for ESRD was LR. 4) No pronounced decrease in renal function after PN was seen during the observation period and most patients who underwent PN in a solitary kidney were able to avoid longterm dialysis, provided that they did not experience LR and subsequent nephrectomy. The results of multiple regression analysis for postoperative GFR changes seem counterintuitive, in that cold ischemia time was associated with immediate GFR decreases, while warm ischemia was not. However, average cold ischemia time was significantly longer than warm ischemia time (33 vs 12 minutes). No patient had a warm ischemia of longer than 25 minutes and no patient with cold ischemia had a clamp time of less than 20 minutes. This is similar to findings in a recent multi-institutional study, in which a greater risk of short-term acute renal failure was also seen in patients with cold vs warm ischemia.4 The limit for safe cold ischemia time in that study was 35 minutes, which was exceeded in half of the patients with cold ischemia in our study. Likewise, in the previous series the stated limit for safe warm time was 20 minutes, which was exceeded by only 2 of 22 study patients with warm ischemia in our study. Our results are consistent with the findings of that study that estimated safe warm and cold ischemia times at 20 and 35 minutes, respectively, although slightly longer warm ischemia time may also be safe. The fact that most patients presenting with tumor in a solitary kidney had baseline renal insuffi-
2042
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
ciency in this series is consistent with the findings of Huang et al, who found only a 30% 3-year probability of maintaining a GFR of greater than 60 ml per minute per 1.73 m2 after RN.2 Fortunately patients with a solitary kidney are able to avoid ESRD after PN and most patients without CKD (GFR greater than 60 ml per minute per 1.73 m2) remained so. Patients with the lowest GFR (less than 35 ml per minute per 1.73 m2) are at higher risk for ESRD but 2 of the 3 in this category who did not experience LR were able to remain dialysis-free for at least 80 months. Therefore, no GFR cutoff should be used to exclude candidates for PN, although patient counseling regarding possible ESRD is important in those with a low GFR. Renal function does not appear to markedly decrease in the intermediate term in patients who do not show ESRD immediately. The single patient with progression to ESRD at 80 months had diabetes with severe cardiovascular disease. He underwent repeat coronary bypass grafting between PN and the development of ESRD. Our results show that the greatest threat to ESRD-free survival is LR. Five of the 13 patients with LR went on to ESRD due to completion nephrectomy and margin status was the factor with the greatest effect on LR. Elective PN series have not shown that positive margins have this same significant effect.8,9 A possible reason for this difference is that lesions in this series were not ideal tumors for PN. Of RCC cases 17 (23%) involved T3 lesions, which is a higher rate than in many other series.5,8,9 Three of these patients had positive margins and recurrence. The series by Fergany et al with a comparable rate of T3 lesions did not describe the effect of positive margins on LR.6 Furthermore, 60% of the tumors in the current series were deep or central. It is likely that positive margins in this study were more significant due to more invasive lesions and a larger tumor-parenchyma interface. In fact, 10 of the 13 cases of positive margins were in
deep or central tumors. It is possible that positive margins in the deepest portion of the tumor bed were more likely to result in LR because this area is usually not fulgurated with an argon beam coagulator to avoid damaging sutures placed for hemostasis or collecting system repair. We did not determine the effect of LR on disease specific survival due to the many patients with previous metastatic disease or large contralateral tumors in our study population, which confounded analysis. This study was limited by a relatively small number of patients and particularly by the limited number of patients who underwent arterial clamping. It is possible that renal ischemia has a long-term effect on GFR that we could not detect at a statistically significant level. Our study has the advantage of being relatively enriched with higher T stage lesions, which allowed us to determine a difference in LR for positive margins in this higher risk population.
CONCLUSIONS PN for tumors in a solitary kidney results in small changes in long-term renal function and good local disease control even in patients with marginal preoperative renal function, multifocal tumors or a history of metastatic disease. However, an effort should be made to limit renal ischemia. Positive margins in solitary kidneys are more likely to be problematic than in elective cases, particularly in those with higher T stage lesions. Every effort should be made to achieve negative margins, including sending frozen sections or briefly using an arterial clamp to improve visualization, since the effects of positive margins appear to be more significant than the longterm effects of brief arterial occlusion. When final pathological results reveal a positive margin, immediate completion nephrectomy may not be justified but patients require close monitoring with a high degree of suspicion for LR.
REFERENCES 1. Rabbani F, Herr HW, Almahmed T and Russo P: Temporal change in the risk of metachronous contralateral renal cell carcinoma: influence of tumor and demographic factors. J Clin Oncol 2002; 20: 2370. 2. Huang WC, Levey AS, Serio AM, Snyder M, Vickers AJ, Raj GV et al: Chronic kidney disease after nephrectomy in patients with renal cortical tumors: a retrospective cohort study. Lancet Oncol 2006; 7: 735. 3. Go AS, Chertow GM, Fan D, McCulloch CE and Hsu C: Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004; 351: 1296.
4. Thompson RH, Frank I, Lohse CM, Saad IR, Fergany A, Zincke H et al: The impact of ischemia time during open nephron-sparing surgery on solitary kidneys: a multi-institutional study. J Urol 2007; 177: 471.
7. Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S et al: Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate. Ann Intern Med 2006; 145: 247.
5. Ghavamian R, Cheville JC, Lohse CM, Weaver AL, Zincke H and Blute ML: Renal cell carcinoma in the solitary kidney: an analysis of complications and outcome after nephron-sparing surgery. J Urol 2002; 166: 454.
8. Kwon EO, Carver BS, Snyder ME and Russo P: Impact of surgical margins in patients undergoing partial nephrectomy for renal cortical tumors. BJU Int 2006; 99: 286.
6. Fergany AF, Saad IR, Woo L and Novick AC: Open partial nephrectomy for tumor in a solitary kidney: experience with 400 cases. J Urol 2006; 175: 1630.
9. Yossepowitch O, Thompson RH, Leibovich BC, Eggener SE, Pettus JA, Kwon ED et al: Positive surgical margins at partial nephrectomy: predictors and oncological outcomes. J Urol 2008; 179: 2158.
NEPHRECTOMY OUTCOMES IN SOLITARY KIDNEYS
2043
EDITORIAL COMMENT Changes in urological practice based on recent information can lessen the likelihood of facing a new tumor in a solitary kidney. PN is equivalent to RN for T1 renal cortical tumors less than 7 cm, of which 20% are benign and 25% are indolent. Most patients with renal tumors are older than healthy kidney donors (median age 61 years) and many have significant comorbidities, including CKD (26%), diabetes, hypertension, a high body mass index and cardiovascular disease. RN is an independent risk factor for CKD and decreased overall survival, and it carries an increased risk of a cardiovascular event after surgery. Despite stage migration to small, incidental renal masses (T1, median less than 4 cm) evidence from national databases indicates that PN remains
underused (less than 20% of T1 cases). The solitary kidney is susceptible to operative damage whether no ischemia, warm ischemia or cold ischemia is used. Early measurements of serum creatinine or the estimated GFR may underestimate long-term ischemic effects. Increased training in open or minimally invasive PN, increased active surveillance in the elderly population and in comorbidly ill patients with small tumors, and the restriction of RN to massive tumors is imperative. Paul Russo Urology Service Department of Surgery Memorial-Sloan Kettering Cancer Center New York, New York
REPLY BY AUTHORS While it is true that some functional loss is typically incurred by the kidney undergoing resection, our results add to the evidence that PN is well tolerated for solitary kidneys and should be considered the gold standard.1,2 Ablative therapies or observation, while certainly appropriate for the elderly or those with significant comorbidities, should not be pursued solely out of concern for functional loss of the solitary kidney. However, our results should introduce a note of caution regarding further expanding indications for PN. Preservation of renal function is clearly preferable3 but we must be careful as we proceed towards PN for larger and more difficult tumors. Some of these cases might be better managed with RN rather than PN if a negative surgical margin cannot be reliably achieved.
One might have the impression from recently published PN series that margin status is not particularly important. However, the majority of lesions previously treated with PN in these series have been small peripheral lesions that are likely to have been slow growing tumors in the short term. Positive margins in this population may not be as predictive of short-term future LR as they would be in larger or more centrally located tumors. For larger or more centrally located lesions PN should be performed with particular care, preferably by those experienced with resection of more difficult tumors, to achieve the same oncological and functional results seen with small peripheral tumors. Cases with a positive margin should be monitored closely with a high degree of suspicion for LR, which may then be salvaged with repeat PN, RN or ablative techniques as necessary.
REFERENCES 1. Ghavamian R, Cheville JC, Lohse CM, Weaver AL, Zincke H and Blute ML: Renal cell carcinoma in the solitary kidney: an analysis of complications and outcome after nephron-sparing surgery. J Urol 2002; 168: 454.
2. Adkins KL, Chang SS, Cookson MS and Smith JA: Partial nephrectomy safely preserves renal function in patients with a solitary kidney. J Urol 2003; 169: 79.
3. Huang WC, Elkin EB, Levey AS, Jang TL and Russo P: Partial nephrectomy versus radical nephrectomy in patients with small renal tumors—is there a difference in mortality and cardiovascular outcomes? J Urol 2009; 181: 55.