The operative safety and oncological outcomes of laparoscopic nephrectomy for T3 renal cell cancer

Authors


Grant D. Stewart, Edinburgh Urological Cancer Group, Department of Urology, University of Edinburgh, Western General Hospital, Crewe Road South, Edinburgh EH4 2XU, UK. e-mail: grant.stewart@ed.ac.uk

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

Laparoscopic radical nephrectomy is a well established treatment for localized RCC, where nephron-sparing approaches are not appropriate. As surgeon and departmental experience grow more extensive tumours will be tackled laparoscopically. However, little is known about the operative safety and oncological outcomes of the laparoscopic approach for locally advanced RCC.

The present study describes the largest reported cohort of patients receiving laparoscopic radical nephrectomy for locally advanced RCC. In the context of suitably experienced personnel in an established centre, we have established that this approach is safe from operative, postoperative and oncological standpoints, with comparable data to existing open series.

OBJECTIVE

  • • To determine the operative, postoperative and oncological outcomes of laparoscopic radical nephrectomy (LRN) for locally advanced renal cell cancer (RCC), which, as surgeon and departmental experience increases, is being performed more often.

PATIENTS AND METHODS

  • • In total, 94 consecutive patients receiving LRN for pathologically confirmed T3 or T4 RCC at a tertiary referral centre between March 2002 and May 2010 were analyzed.
  • • Preoperative, operative, tumour and postoperative characteristics were evaluated together with recurrence and outcome data.
  • • Survival was estimated using the Kaplan–Meier method. Cox's proportional hazards model was used for multivariate analysis.

RESULTS

  • • In total, 77 patients had LRN with curative intent and 17 patients had LRN with cytoreductive intent.
  • • There were six LRNs (6.4%) that were converted to open procedures.
  • • Overall, there were two (2.1%) Clavien grade IIIa complications, one (1.1%) grade IVa complication and one (1.1%) postoperative death.
  • • Overall median follow-up was 17.4 months. In total, 22 (28.6%) patients receiving curative LRN developed a recurrence after a median of 13.9 months; 12 (54.5%) patients developed distant metastases, five (22.7%) patients had local recurrences and three (13.6%) patients had transcoelomic spread. Median predicted progression free survival was 48.4 months in patients undergoing LRN with curative intent. Median predicted overall survival was 65.6 months after curative LRN and 15.7 months after cytoreductive LRN.
  • • Multivariate analysis did not reveal any variables influencing recurrence or survival.

CONCLUSIONS

  • • In the context of suitably experienced personnel in an established centre, LRN for locally advanced RCC is safe from an operative and oncological standpoint.
  • • Patients clinically staged as T3 RCC must still be selected carefully for LRN in a multidisciplinary setting.
Abbreviations
AJCC

American Joint Committee on Cancer

CSS

cancer-specific survival

LRN

laparoscopic radical nephrectomy

OS

overall survival

PFS

progression-free survival

UICC

Union Internationale Contre le Cancer.

INTRODUCTION

Extirpative surgery is the only curative treatment for RCC. Where partial nephrectomy is not amenable or suitable, it is now well established that laparoscopic radical nephrectomy (LRN) represents a first-line treatment for localized RCC [1]. There is substantial evidence showing that the operative, postoperative and oncological outcomes of LRN for T2 RCC are at least equivalent to those for open nephrectomy [2–4]. As experience with LRN grows, case selection has expanded to include more complex cases, which means that carefully selected locally advanced RCC (i.e. T3 and T4 disease) will be operated on laparoscopically. Furthermore, larger tumours that are clinically/radiologically staged as T2 RCC may be understaged and only identified as T3 RCC at the time of pathological analysis [5]. The European Association of Urology guidelines state that the treatment of cT3 RCC should comprise radical nephrectomy and that LRN is feasible in selected patients [1]. However, reported data on the outcomes of LRN for locally advanced RCC [6–9] are scarce. The present study aimed to determine the operative, postoperative and oncological outcomes of LRN for locally advanced RCC performed in a single unit with experience of 850 laparoscopic nephrectomies over a 19-year period.

PATIENTS AND METHODS

In total, 94 consecutive patients diagnosed with pT3 and pT4 RCC who underwent LRN in our unit from March 2002 to May 2010 were analyzed. Union Internationale Contre le Cancer (UICC)/American Joint Committee on Cancer (AJCC) TNM sixth edition staging for RCC was used to stage the tumours [10]. All cases and their radiological imaging were discussed at a multidisciplinary meeting and the operations were performed by one of three surgeons using a standard transperitoneal approach (D.A.T., S.A.M. and A.C.P.R.). Of note, the kidney was extracted in an impermeable bag or via open removal.

Hospital medical records were reviewed retrospectively to assess preoperative CT staging, postoperative course, pathology, follow-up, disease recurrence, death and cause of death. If causes of death were not available in clinical notes, they were obtained from death certificates from the General Registry Office of Scotland. Postoperative complications were recorded using the Clavien–Dindo classification [11]. A standardized follow-up protocol is used in our department (and remained similar across the entire time period of the study) and consists of history, examination, blood tests, chest radiograph and ultrasonography or a CT scan if symptoms, signs or investigations warrant this. Follow-up took place at 3 months after surgery, then 6-monthly for 3 years, then annually for a further 2 years and, subsequently, every 5 years if the patient was aged < 60 years. A chest and abdominal CT scan was routinely performed at the 2-year time-point after surgery and before discharge at 5 years. The last follow-up date was considered as the last time that the patient was seen and confirmed to be free of recurrence. Length of follow-up was determined by the time between surgery and most recent clinical contact. Local disease recurrence was defined as RCC recurrence in the renal bed or inferior vena cava. Transcoelomic spread was defined as peritoneal or bowel recurrence. Distant spread was metastasis to a distant solid organ or lymph node chain. RCC development in the contralateral kidney was not included as recurrence after the original surgery. From the date of disease recurrence, the duration of progression-free survival (PFS) could be calculated. Cancer-specific survival (CSS) was determined if the cause of death was directly attributable to RCC according to clinical notes or death certificates.

Continuous variables were compared using an unpaired t-test and categorical variables were analyzed using a chi-squared test or Fisher's exact test. PFS, CSS and overall survival (OS) were estimated using the Kaplan–Meier method. Survival multivariate analysis was performed using the Cox's proportional hazards model, which was calculated to be powered adequately for use with up to eight variables. Statistical analysis was performed using PASW, version 18.0 (SPSS Inc., Chicago, IL, USA). P < 0.05 was considered statistically significant.

RESULTS

The number of LRNs for locally advanced RCC increased with the cumulative experience of the unit, which had carried out 846 laparoscopic nephrectomies by the time of the present analysis (Fig. 1).

Figure 1.

Bar chart showing the number of laparoscopic radical nephrectomies performed for locally advanced disease in each consecutive year. The cumulative numbers of laparoscopic nephrectomies, for all pathologies, are denoted in parentheses. Note that the results shown for 2010 refer to the period from January to May only.

Of the 94 patients who had LRN for pT3 or pT4 RCC, 77 (81.9%) patients had LRN with curative intent and 17 (18.1%) patients had LRN for metastatic RCC with cytoreductive intent. Table 1 provides the patient, tumour, operative and postoperative characteristics of the patients included in the analysis. Other than tumour grade, which was higher in the cytoreductive patients, there were no differences between the characteristics of the curative and cytoreductive groups. Of the 72 patients for whom data were available, 61 (84.7%) patients had their kidney extracted from the abdominal cavity using an impermeable bag. The remaining 11 patients had their kidney extracted via an open technique; of these patients, six had had a converted LRN, two patients had a laparoscopic nephroureterectomy with open resection of distal ureter because their tumour was considered to be a TCC, one patient had a combined open hysterectomy, and there was no explanation found for the remaining two patients. None of the patients undergoing an open kidney extraction developed a local or transcoelomic recurrence.

Table 1.  Descriptive characteristics of 94 patients treated with laparoscopic radical nephrectomy (LRN) for pT3/T4 between March 2002 and May 2010 for both curative intent and cytoreductive for metastatic RCC
VariableCurative intent (n = 77)Cytoreductive (n = 17) P
  1. *Patient staged as a cT2 RCC and had an attempted LRN, which was converted to open surgery because of failure to progress. This patient went on to develop renal bed, peritoneal, lung and colonic recurrence; required a colonic stent for bowel obstruction 3 months after nephrectomy; and died 4.5 months after surgery. †Patient had identifiable lung metastases but no renal vein involvement on staging CT; at LRN, the kidney was abutting the liver but could be resected laparoscopically. This patient died from metastatic RCC 15.7 months after surgery subsequent to adjuvant sunitinib therapy. ‡Two conversions were for failure to progress; two conversion were a result of no plane between the liver and upper pole of kidney; and two conversions were for haemorrhage (one when a clip was dislodged from the renal artery whilst putting the kidney into a bag). §Chest drain for pneumothorax. ¶Percutaneous drainage of collection. **Anaphylactic reaction to gelofusin and was admitted to an intensive treatwent unit for therapy. ††A man aged 82 years with Parkinson's disease who had a combined LRN and laparoscopic colectomy and died from respiratory failure.

Patient characteristics:   
 Age (years), Median (range)68.0 (36.1–84.9)63.6 (42.9–74.2)0.26
 Sex, n (%)  0.79
  Male48 (62.3)10 (58.8) 
  Female29 (37.7)7 (41.2) 
Tumour characteristics:   
 Laterality, n (%)  0.54
  Right38 (49.4)9 (41.2) 
  Left39 (50.9)10 (58.8) 
 Tumour size (maximum diameter, cm), Median (range)7.0 (2.5–17.0)9.0 (4.0–18.7)0.15
 Grade, n (%)  0.010
  11 (1.3)0 
  242 (54.5)2 (11.8) 
  321 (27.3)8 (47.1) 
  413 (16.9)7 (41.2) 
 Stage, n (%)  0.44
  pT3a28 (36.4)7 (41.2) 
  pT3b48 (62.3)9 (52.9) 
  pT41 (1.3)*1 (5.9) 
 Histological subtype, n (%)  0.65
  Clear cell67 (87.0)15 (88.2) 
  Papillary5 (6.5)2 (11.8) 
  Chromophobe3 (3.9)0 
  Translocation tumours2 (2.6)0 
Operative characteristic:   
 Laparoscopic conversions, n (%)6 (7.8)00.59
 Estimated blood loss (mL)   
  Median (range)100 (0–3000)200 (10–2600)0.47
 Operation duration (min)   
  Median (range)150 (80–320)145 (127–260)0.53
Postoperative complications, n (%):  0.20
 No complications30 (46.9)6 (40.0) 
 Clavien grade I15 (23.4)5 (33.3) 
 Clavien grade II17 (26.6)2 (13.3) 
 Clavien grade IIIa1 (1.6)§1 (6.7) 
 Clavien grade IIIb00 
 Clavien grade IVa01 (6.7)** 
 Clavien grade IVb00 
 Clavien grade V (death)1 (1.6)††0 
Postoperative characteristics:   
 Length of postoperative stay (days), Median (range)5 (3–81)6 (3–11)0.53
 Length of follow-up of censored patients (months)  0.67
  Mean (median)24.8 (17.4)22.6 (19.5) 
  Range0.2–98.116.9–34.3 

Of the 77 patients who were operated on with curative intent, 48 were found to have pT3b RCC. Although all patients had a preoperative CT scan, the details of these scans were available for 37 (77.1%) pT3b patients. Of these 37 patients, eight (21.6%) were noted to have renal vein involvement on CT, suggesting upstaging on final pathology for the remaining 29 (78.4%) patients. Those patients for whom renal vein involvement was noted on CT scan were generally considered to be amenable to LRN because the tumour thrombus extended only to the proximal vein.

In total, 22 (28.6%) patients with non-metastatic RCC at the time of LRN subsequently developed recurrence after a median (range) of 13.9 (1.0–68.3) months. There were 12 (15.6%) patients who developed a distant recurrence (Table 2). Table 3 describes the further management of the patients with localized RCC who subsequently developed a recurrence, as well as the 17 patients receiving cytoreductive LRN.

Table 2.  Detailed description of the 22 (28.6%) patients with non-metastatic RCC at the time of laparoscopic radical nephrectomy who subsequently developed disease recurrence
Recurrence sitePatients,n (%)
  1. *One patient to mediastinal lymph nodes, abdominal lymph nodes and liver; one patient to lungs, spleen, and mediastinal lymph nodes; one patient to lung, contralateral kidney and adrenal; one patient to lung and breast; and one patient to pancreas, liver and right iliac fossa. †One patient to the inferior vena cava; two patients to renal bed and peritoneum; and two patients to renal bed. ‡Two patients to peritoneum; and one patient to small and large bowel.

Liver1 (4.5)
Lungs3 (13.6)
Bone1 (4.5)
Mediastinal lymph nodes2 (9.1)
Multiple distant sites*5 (22.7)
Local recurrence5 (22.7)
Transcoelomic spread3 (13.6)
Unknown site (died from metastatic RCC)2 (9.1)
Table 3.  Details of the management of the 22 patients with non-metastatic RCC who subsequently developed a recurrence and the 17 patients receiving cytoreductive laparoscopic radical nephrectomy
TherapyCurative patients who developed recurrence (n = 22), n (%)Cytoreductive (n = 17), n (%)
  1. *Both as part of the MRC RE04/EORTC GU 30012 trial [24], one patient had monotherapy and one had triple therapy. †One patient treated with interferon and then sunitinib; and one patient had interferon as part of the MRC RE04/EORTC GU 30012 trial [24].

Immunotherapy2 (9.1)*2 (11.8)
Sunitinib3 (13.6)7 (41.2)
Everolimus1 (4.5)0
Surgery3 (13.6)0
Observation5 (22.7)3 (17.6)
Deterioration in performance status precluding treatment8 (36.4)5 (29.4)

Table 4 provides the estimated median/mean survival and 5-year survival for the study cohort. The overall median follow-up for all patients was 17.4 months. As expected, there were significant differences in OS, CSS and PFS between patients receiving cytoreductive vs curative LRN for pT3/4 RCC (P < 0.001; log-rank test for all survival analyses) (Fig. 2). There were no significant differences for OS (P = 0.82), CSS (P = 0.67) or PFS (P = 0.35) between pT3a and pT3b RCC in patients who were operated on with curative intent, or when curative and cytoreductive cases were considered together (OS, P = 0.80; CSS, P = 0.60; PFS, P = 0.99). Considering all patients together (i.e. cytoreductive and curative intent), there was a significant difference in predicted OS (P = 0.006; log-rank test), predicted CSS (P < 0.001) and predicted PFS (P < 0.001) when comparing low grade (1 + 2) and high grade (3 + 4). When considering LRN for non-metastatic disease, there was no significant difference in OS (P = 0.64) or CSS (P = 0.12) when comparing low and high grade; however, there was a significant difference with respect to PFS (P = 0.024) (Fig. 3). The histological subtype (i.e. clear cell RCC vs non-clear cell RCC) did not influence predicted OS (P = 0.92), CSS (P = 0.64) or PFS (P = 0.10).

Table 4.  Estimated mean/median survival and 5-year survival for overall survival (OS), progression-free survival (PFS) and cancer-specific survival (CSS) for laparoscopic radical nephrectomy with curative and cytoreductive intent for T3/4 RCC using the Kaplan–Meier method
FactorAnalysisMean survival, months (years)Median survival, months (years)5-year survival (%)
  1. NA, not available.

OSOverall54.3 (4.5)46.1 (3.8)43.7
Curative intent63.9 (5.3)65.6 (5.5)54.2
Cytoreductive16.4 (1.4)15.7 (1.3)0
PFSCurative intent46.5 (3.9)48.4 (4.0)44.6
CSSOverall61.8 (4.8)NA51.1
Curative intent72.5 (6.0)NA62.6
Cytoreductive17.1 (1.4)15.7 (1.3)0
Figure 2.

Overall survival (A), cancer-specific survival (B) and progression-free survival (C) after laparoscopic radical nephrectomy for pT3/4 RCC with either curative or cytoreductive intent, estimated using the Kaplan–Meier method. There were significant differences were found for all these comparisons (P < 0.001; log-rank test for each comparison).

Figure 3.

When comparing low grade (1 + 2) and high grade (3 + 4) considering laparoscopic radical nephrectomy with curative intent, there was a significant difference for progression-free survival (P = 0.024; log-rank test).

When sex, age, tumour size, histological subtype, tumour stage and tumour grade were used as variables in a Cox's proportional hazards model, none of the variables were in the equation for PFS, CSS and OS.

DISCUSSION

The present study has shown that LRN for locally advanced RCC can be performed safely in the context of experienced laparoscopic urological surgeons in a department with significant experience of the procedure. The operative, postoperative and oncological outcomes are all equivalent to open nephrectomy performed for locally advanced RCC [12–14]. Considering the cases undergoing surgery, the results obtained in the present study show an acceptable level of conversion to open operation, blood loss, operation duration, postoperative complications and postoperative mortality [15,16]. Concerns over high levels of local recurrence are not supported by the present series.

There are a number of strengths to the present study, which describes the largest reported series of patients with pT3/4 RCC treated by LRN. First, the present study comprises a single-centre study, which adds uniformity to the follow-up and the results attained. Second, three surgeons performed all of the procedures included in the present study, they performed LRN using a standard approach and they sequentially mentored each other. Third, much of the data presented were collected prospectively with retrospective validation. The potential limitations of the present study are also recognized. Importantly, the present study only assessed the outcomes of pathologically-staged T3 RCC rather than clinically-staged disease. Additionally, the exact extent of any renal vein tumour thrombus was not clear in every case. Furthermore, the median follow-up of 17.4 months was relatively short, although the median time to progression of T3 RCC has previously been reported as 17 months [12]. Finally, we cannot be explicit about the exact selection criteria used for the patients for LRN rather than open surgery, although it is clear from the data that the indications expanded as departmental experience increased. For example, with increasing experience, larger tumours were approached laparoscopically, as were tumours with limited extension into the renal vein on preoperative imaging.

The operative characteristics (estimated blood loss and operation duration) of patients undergoing LRN for locally advanced RCC were similar to the existing literature reporting on LRN for pT3 RCC [7]. The rate of conversions in the present study fell within the acceptable range for LRN of 0–10% [17]. There were only three (3.1%) patients in the whole series who required intervention for postoperative complications and one postoperative death; this also fits into the range previously reported for LRN [17]. The median postoperative stays of 5 days and 6 days, respectively, for the curative and cytoreductive groups is slightly longer than the published ranges for LRN for localized disease (3.2–4.5 days) but shorter than most open nephrectomy series (5.1–8.9 days) [18]. Accordingly, from both technical and postoperative standpoints, it is safe to perform a LRN for pT3 RCC.

Interestingly, only 21% of the patients with pT3b RCC were identified as having renal vein invasion on their preoperative cross-sectional imaging. Segmental renal vein involvement, as opposed to main renal vein involvement, may not easily be identified on cross-sectional staging. As such, most patients in the present study were not known to have locally advanced primary tumours before their surgery.

Any concerns that LRN for aggressive RCC could result in a high level of local recurrence are not supported by the results of the present study. 6.5% patients developed a local recurrence compared to a 8.3–9.3% local recurrence reported in previous open series [12,19]. However, two patients developed peritoneal recurrences and one patient developed an unusual small and large bowel recurrence, probably as a result of direct transcoelomic spread. There are only case reports available for previous examples of RCC metastasizing to the small bowel [20]. The rate of transcoelomic spread (3.9% in the present series) has not been widely reported in the literature. However, there have been descriptions in the literature of peritoneal recurrence after nephrectomy for RCC [21], although it is not possible to determine whether these transcoelomic recurrence events were potentiated by the aggressive nature of the RCC or the use of a transperitoneal laparoscopic approach. In a study investigating 44 patients with pT3 RCC, Bensalah et al. [8] did not report any peritoneal recurrences, although the mean tumour size in their series was considerably smaller (5 cm) than the series reported above [8]. Of note, in the study presented here, there were no port site metastases, which is another theoretical concern with LRN [22], despite aggressive RCCs being treated.

Previous open nephrectomy studies describe an incidence of RCC metastasis of 39% for pT3 tumours and a median PFS of 17 months [12]. Although not directly comparable, the results of the present study show a superior outcome after LRN for non-metastatic pT3/4 RCC, with a metastasis rate of 29% and a median predicted PFS of 48 months. The predicted 5-year PFS is 44%. It is possible that the superior PFS results after LRN are related to patient selection bias. There are few descriptions of oncological outcomes after LRN for pT3 RCC. Bensalah et al. [8] reported that only 7% of patients receiving LRN for non-metastatic pT3 RCC went on to develop metastatic RCC and die from the disease [8]; this is compared with 11 (14.3%) patients operated on with curative intent in the present series. An explanation for the lower mortality rate reported by Bensalah et al. [8] may be the reported inclusion biases, meaning that the T3 tumours described may not reflect the outcome of typical clinical T3 tumours (because, amongst other factors, they were smaller compared to other series) [8]. Finally, a CSS of 37–44% for pT3 RCC and 20–28% for pT4 RCC has been shown in larger studies (although no details are provided regarding whether these procedures were open, laparoscopic or a mixture of both approaches) [13,14]. The overall results of the present study (i.e. a 5-year CSS of 51%) compare favourably with these previous studies.

The median predicted OS of the entire cohort in the present study was 3.8 years, with an anticipated significant difference in median survival (P < 0.001; log-rank test) between patients operated on with curative intent (5.5 years) and those treated with a cytoreductive nephrectomy (1.3 years). These values, together with the 5-year predicted OS of 54% and CSS of 63% for curative patients, are very helpful for counselling patients after their surgery and explaining the need for very close follow-up. The predicted 5-year OS and CSS was 0% for patients undergoing a cytoreductive nephrectomy.

Classifying patients using the UICC/AJCC TNM sixth edition relative to the seventh edition is useful because it allows the delineation of patients with pT3 disease into those with perirenal/renal sinus fat invasion (pT3a) and those with segmental or main renal vein invasion (pT3b). Renal vein thrombus can often be dealt with laparoscopically by milking back the tumour (often occurring spontaneously after ligation of the renal artery) and division of the vein distal to the tumour tongue. By contrast, segmental renal vein involvement may be unrecognized at the time of surgery and appears to be little different from LRN for T2 RCC. Using the UICC/AJCC TNM seventh edition, all of the pT3 patients in the present study would be staged as pT3a (perirenal/renal sinus fat invasion, as well as segmental and main renal vein tumour thrombus) because none of these patients had tumour extension into the inferior vena cava [23].

In conclusion, we have described the largest reported cohort of patients receiving LRN for locally advanced RCC, which represents an approach for which very few series have been published [6–9]. This approach is safe from operative, postoperative and oncological standpoints. Furthermore, useful survival data are provided for the counselling of patients with locally advanced RCC.

CONFLICT OF INTEREST

None declared.

Ancillary