SEARCH

SEARCH BY CITATION

Keywords:

  • kidney;
  • nephrectomy;
  • local recurrence;
  • neoplasm

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

The local recurrence of renal cancer is an uncommon event, but it is often tempting to remove such recurrences surgically. Authors from the USA analysed the survival benefit of such a strategy in the presence and the absence of concomitant metastatic disease. They found that, in the absence of metastases, complete surgical resection of local recurrences is associated with improved survival.

Authors from the same institution compared outcomes in elective partial vs radical nephrectomy for renal cancers of 4.7 cm. They found no clear evidence that partial nephrectomy was associated with worse cancer control, but suggest that continued follow-up of this cohort is required.

OBJECTIVE

To analyse the survival benefit of resecting local recurrence (LR) of renal cell carcinoma (RCC) in the presence and absence of concomitant metastasis.

PATIENTS AND METHODS

From 1989 to 2004 we identified 34 patients with LRs (2.9%) of the 1165 radical nephrectomies performed for T1–4N0M0 disease. Of these, 18 (53%) had no evidence of metastasis (isolated LR incidence, 1.5%) and 16 (47%) had synchronous metastasis. Of the 18 patients with no metastasis, 11 had complete surgical resection (group I) and seven had nonsurgical therapy (group II). Of the 16 patients with synchronous metastasis, five had surgery (group III) and 11 did not (group IV). Survival was projected using the Kaplan–Meier method and log-rank test for each group.

RESULTS

Eight of the 34 patients (24%) were symptomatic. The T stage of the initial nephrectomy was T1a in two cases, T1b in six, T2 in five, T3a in six, T3b in eight, T4 in six and unknown in one; 22 patients (65%) had clear cell histology. There were no significant differences in median time to LR or the LR size among the groups. The median (range) follow-up was 16.9 (0.5–103.6) months. Of the 11 patients in group I, three remain with no evidence of disease, three are alive with metastatic disease, and five died from disease. By contrast, 21 of the 23 patients (91%) in groups II, III and IV died from disease. The overall estimated 1-, 3- and 5-year survivals were 63%, 31% and 18%. The median survival time was 71.4 months for group I, 9.9 for II, 16.3 for III, and 11.8 for IV (P < 0.01) with a 5-year survival of 62% for group I and 0% for groups II, III, and IV.

CONCLUSIONS

LR after radical nephrectomy is rare (2.9%) and has a poor prognosis. The presence of synchronous metastasis and nonoperative therapy are related to these low survival rates. However, if there is no metastatic disease, complete surgical resection of LRs is associated with improved survival.


Abbreviations
LR

local recurrence

DVT

deep vein thrombosis

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

After radical nephrectomy, local recurrence (LR) of RCC is rare, occurring in only ≈ 2% of patients, and isolated LR with no evidence of metastatic disease is even rarer. Because of this low incidence, many treatment methods have been attempted, none of which has become standard. For isolated LRs, radiation therapy and immunotherapy have been largely supplanted by surgical resection alone. Even so, these patients are at high risk of developing distant metastasis. In the last decade, several studies promoted surgical resection of isolated LRs as the only way to improve survival [1–3]. Improved imaging techniques combined with regimented follow-up protocols might allow a LR to be diagnosed earlier, before symptoms or potentially metastatic disease are apparent.

A corollary to this management is the treatment of LRs in the presence of distant metastases (i.e. lung, bone, etc.). Does surgical resection of the LR with concomitant metastasectomy improve survival, or do these patients have a similar poor survival to those who are not operated on? To answer these questions, we analysed the survival benefit of LR resection with and without concomitant metastasis.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

From our prospective departmental database we identified 34 patients from 1989 to 2004 who were evaluated for LR after radical nephrectomy for RCC. In this 15-year period, 1165 radical nephrectomies were performed at our institution for T1–4N0M0 disease, yielding a LR incidence of 2.9%. All 34 patients had had an open unilateral radical nephrectomy and regional lymphadenectomy, with a pathologically confirmed diagnosis of RCC.

LR was defined as disease development in the renal fossa, ipsilateral adrenal gland or ipsilateral retroperitoneal lymph nodes. LRs were determined by CT of the abdomen, and symptoms were evaluated with serial history and physical examinations. A chest X-ray and/or chest CT as well as a bone scan were used for complete re-staging.

Of the 34 patients, 18 (53%) showed no evidence of metastasis, giving an isolated LR incidence of 1.5%. The other 16 patients (47%) had one or more synchronous metastasis as well as their LR. The 34 patients were stratified into four groups (Fig. 1). Of the 18 patients without metastasis, 11 had complete surgical resection (group I), and seven had nonsurgical therapy (group II). The 16 patients with synchronous metastasis included five who had local resection and metastasectomy (group III), and 11 who had no surgical intervention (group IV). Survival was projected using the Kaplan–Meier method and log-rank test for each group.

image

Figure 1. Patient grouping based on metastasis status and surgical intervention.

Download figure to PowerPoint

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

The median (range) age of the 34 patients at LR was 63 (27–84) years, with equal numbers of men (17) and women (17). Fifteen of the LRs were on the right side and 19 on the left. Twenty-six patients (76%) were asymptomatic; the LR was found on routine imaging. Eight patients (24%) had symptoms, including flank/abdominal pain, haematuria and weight loss. The median (range) time from nephrectomy to diagnosis of LR was 17.0 (1.1–121.7) months.

Using the TNM staging system [4], the initial nephrectomy specimens were staged as: two T1a, five T1b, five T2, six T3a, eight T3b, and six T4; one was unavailable due to inability to obtain an outside record. All cases were N0M0 at initial nephrectomy. Histological subtype classification showed conventional clear cell in 22 patients (65%), chromophobe in four (12%), papillary in one (3%), and other (collecting duct, unclassified, etc.) in seven (20%; Table 1).

Table 1.  The pathology, diagnosis and outcome of patients with a LR
Group/patient no.T StageHistologyTime to LR, monthsSymptomsSite of recurrenceStatusFollow-up, months
  1. CLR, clear cell (conventional) carcinoma; CHR, chromophobe; PAP, papillary; UNC, unclassified; IVC, inferior vena cava; CDT, colleting duct; RPLN, retroperitoneal lymph nodes; NED, no evidence of disease; AWD, alive with disease; DOD, dead of disease; NA, not available.

I
  1T1aCHR 21.1NoIpsilateral RPLNAWD 13.6
  2T1aCLR  9.2NoIpsilateral RPLNDOD 71.4
  3T1bCLR 18.1NoIpsilateral adrenal + IVCDOD 73.5
  4T1bCLR 15.0NoIpsilateral adrenalDOD 56.5
  5T1bCLR103.0YesRenal fossa + psoasNED 38.8
  6T2CHR 76.9NoRenal fossaNED 30.3
  7T2NA 121.7YesRenal fossa + liver invasionAWD  8.2
  8T3aCLR  4.8NoRenal fossaDOD 19.4
  9T3bCLR  8.5NoRenal fossaNED 61.2
 10T3bNA 23.8YesRenal fossaAWD103.6
 11NACDT 10.1NoIpsilateral adrenalDOD 17.7
II
 12T2UNC 28.6NoRenal fossaDOD 33.5
 13T3aCLR 23.4NoRenal fossaAWD  0.5
 14T3aUNC  7.9YesRenal fossaDOD  5.4
 15T3bUNC 18.7NoIpsilateral RPLNDOD  9.9
 16T4CLR 26.5NoRenal fossaDOD  1.0
 17T4CLR  1.1NoRenal fossaDOD  11.8
 18T4CLR  5.0NoRenal fossaDOD 31.6
III
 19T1bCHR 19.2NoIpsilateral RPLNDOD 52.2
 20T1bCLR 22.7YesRenal fossaDOD  3.2
 21T1bCLR 30.0NoRenal fossaDOD 10.4
 22T2CLR 14.0YesRenal fossaDOD 21.2
 23T3bCLR  8.0YesRenal fossaDOD 16.3
IV
 24T2CLR 19.6NoRenal fossaDOD 22.3
 25T3aCLR 17.6NoRenal fossaAWD 13.8
 26T3aCLR  6.0YesRenal fossaDOD  11.8
 27T3aPAP  8.6NoRenal fossaDOD  1.3
 28T3bCHR 16.4NoIpsilateral adrenalDOD 21.7
 29T3bCLR  3.4NoRenal fossaDOD  8.2
 30T3bCLR 80.5NoRenal fossaDOD 47.4
 31T3bCLR 27.2NoRenal fossaDOD  7.3
 32T4CLR 15.1NoRenal fossaDOD 21.8
 33T4CLR  4.9NoRenal fossa + IVCDOD 10.0
 34T4UNC  4.8NoRenal fossaDOD 10.1

Sixteen patients had surgery for LR. Of these, 11 (group I) had no evidence of metastasis on preoperative evaluation and had complete surgical resection of the LR. Six patients managed surgically had a LR in the renal fossa; in one there was invasion into the ipsilateral psoas muscle and in another, invasion into the liver. Three patients had an ipsilateral adrenal recurrence, one of which invaded the wall of the inferior vena cava, and two had ipsilateral retroperitoneal lymph node recurrence. All patients had complete radical resection of the LR with negative surgical margins on final pathology. The mean (range) estimated blood loss was 680 (100–1200) mL and the median (range) hospital stay was 7 (4–16) days. The only perioperative complication was one deep vein thrombosis (DVT) treated with anticoagulation.

Five patients (group III) had a LR and a focus of metastatic disease (three pulmonary, one spine, one femur). These patients had a surgical combination of LR resection and metastasectomy. The mean estimated blood loss was 410 (100–750) mL and the median hospital stay was 7 (4–74) days. There were no postoperative deaths. One patient developed a pneumonia, chylothorax and UTI after partial lung wedge resection. The total perioperative morbidity for groups I and III was 13%.

Eighteen patients had no surgical intervention for their LR. Seven patients (group II) had no evidence of metastasis, but did not have surgery because of radiological evidence of unresectability, poor performance status or patient preference. Eleven patients (group IV) had evidence of metastasis as well as their synchronous LR and were treated as patients with widely metastatic disease. All nonsurgical therapy (including immunotherapy with interferon or interleukin) was according to the treating oncologist's protocol.

The median (range) LR size was 3.8  (1.7–15.0) cm in greatest diameter, and the median follow-up was 16.9 (0.5–103.6) months, with no significant differences between the groups.

Of the 11 patients in group I, three had no evidence of disease at 30.3, 38.8 and 61.2 months, three are alive with metastatic disease and five died from disease (Table 1). By contrast, 21 of the 23 patients (91%) in groups II, III and IV died from disease at a median of 11.8 months. The overall aggregate 1-, 3- and 5-year survivals were 63%, 31% and 18%. The 5-year disease-specific survival of group I was 62%, and was 0% for groups II, III, and IV (Fig 2). The median (95% CI) disease-specific survival time for group I was 71.4 (42.3–100.5) months and for group II was 9.9 (2.1–17.7) months (P < 0.001; Fig. 2A). The median disease-specific survival time for group III was 16.3 (3.6–29.0) months and for group IV was 11.8 (0.2–23.4) months (P = 0.68; Fig. 2B).

image

Figure 2. Disease-specific survival: A, groups I and II (P < 0.001); B, groups III and IV (P = 0.68).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

The incidence of LR for RCC after radical nephrectomy is low. Although the true incidence is unknown, older series reported that over a third of patients have radiographic evidence of LR after nephrectomy on CT [5]. In 1984, Parienty et al.[6] described a 41% incidence of LR on autopsy report of patients who died from metastatic RCC. These high incidences might represent clinically insignificant renal fossa entities overshadowed by the gross metastatic disease. The present series had a much lower incidence of LR (2.9%, 34/1165 patients), probably reflecting clinical and radiographic limitations in identifying small, possibly inconsequential LRs. Isolated LRs with no evidence of metastatic disease are even rarer. Itano et al.[2] reported an isolated LR incidence of 1.8% (30/1737 patients) and Schrodter et al.[3] an incidence of 0.8% (8/1031 patients).

The present data confirm this low incidence; 18 of 1165 patients had isolated LR (1.5%). These patients initially presented with a LR at a median of 17 months. Although this includes cases with and without concomitant metastasis, we found no significant differences between these two groups. The time to LR is similar to that found by Tanguay et al. [1] of 16.5 months and Itano et al.[2] of 18.0 months (Table 2), supporting the use of routine radiographic imaging of the abdomen after nephrectomy. Further supporting this is the large number of patients who were asymptomatic; only 24% of patients in the present series had symptoms indicating relapse of local disease. Similar small percentages were reported by Tanguay et al.[1] (38%) and Schrodter et al.[3] (15%; Table 2).

Table 2.  A review of previous publications on LR
FactorReference
Tanguay et al.[1]Itano et al.[2]Schrodter et al.[3]Present study
  • *

    group I+II,

  • ‡group I, NA, not available.

Year published1996200020022006
Number of patients  16  30  16  34
Median age, years  53  67  62  63
Symptomatic LR, %  38  60  15  24
Time to LR, months  16.5  33.6  45.5  24.2
Patients with isolated LR, n  15  30  13  18*
Patients having surgical resection, n  15  10  13  11
Median survival, monthsNANANA  71.4
5-year survival, %NA  51NA  62

Several recent studies provided evidence that TNM stage corresponds with relapse rates after nephrectomy, and follow-up guidelines should be tailored accordingly [7–9]. Specifically, patients with T3 or T4 disease need more frequent radiographic evaluation than patients with T1 or T2 disease. Although this is more feasible and cost-effective on a large population level, it fails to detect the uncommon, but potentially curable, isolated LR. Indeed, 38% of the present patients had T1/2 disease at initial nephrectomy. Finally, the present data do not indicate that histological subtype is correlated with LR or survival, although this might be due to the small sample size.

Historically, LR even without metastasis has been thought to be a poor prognostic sign. Several studies reported varying survival outcomes for LRs; deKernion et al.[10] reported on 14 patients with LRs and metastatic disease; they had a 1-year survival rate of 14%, vs 40% for patients with metastasis only. Although this appeared as a negative prognostic factor, most of those patients had N1 disease at initial nephrectomy, indicating regional metastasis. In 1992, Esrig et al.[11] reported a 3-year survival rate of 36% for patients with isolated LR who underwent resection. In 1996, Tanguay et al.[1] reported on 16 patients who had surgical resection of an isolated LR with or without biological therapy; 12 of these were alive at ≈ 2 years after resection, and a possible improvement in survival was noted for those who had neoadjuvant immunotherapy.

In 2000, Itano et al.[2] reported on 30 patients with an isolated LR treated by surgery, adjuvant medical therapy or observation. The overall 5-year disease-specific survival was 28%, but those who had surgery had a 51% 5-year survival rate, vs 18% for those treated with medical therapy, and 13% for those managed by observation. In 2002, Schrodter et al.[3] reported on 13 cases of isolated LR treated by surgical extirpation; of these, seven died from disease after a mean of 23.1 months, one was alive with disease and five showed no evidence of disease with a mean survival time of 52.4 months.

In the present study, the overall 1-, 3-, and 5-year survival rates were 63%, 31% and 18% for patients with or without metastasis. As expected, group I (no metastasis + surgery) had the best survival rates; of 11 patients, three show no evidence of disease, three are alive with disease and five died from disease. Although only three patients are clinically ‘cured’, the survival curves indicates a substantially longer life-span, with a 5-year disease-specific survival rate of 62% and a 5-year disease-free survival of 51%. The median survival time for this group was ≈ 6 years. These data, which are similar to those reported by Itano et al.[2] (Table 2), supports an aggressive surgical approach to patients with no evidence of disease.

By contrast to group I, the other groups fared poorly; seven patients with isolated LRs who did not have surgical resection (group II) had a median survival of only 9.9 months, and all but one died from disease. Five patients who had surgical resection of all disease for the LR and metastatic focus (group III) also did poorly, with a median survival of 16.3 months; all five were dead from disease at 5 years. Finally, of 11 patients with a LR and metastasis who had no surgical intervention (group IV) 10 died from disease within a median of 11.8 months.

The present data show that surgical extirpation gives a significant survival advantage to patients with an isolated LR. Although this surgery is not without complications, we found a low morbidity (13%) including a DVT, pneumonia and UTI, but no perioperative mortality. In comparison, Tanguay et al.[1] reported a 31% complication rate (subphrenic abscess, pneumonia, pyelonephritis) and Itano et al.[2] a 33% rate (ileus, pneumonia, pneumothorax).

These findings suggest that every attempt should be made to offer curative surgical intervention to patients with an isolated LR. However, patients with a LR and synchronous metastatic disease will not benefit from surgical resection, and should be offered systemic treatment. The development of new and possibly more effective systemic treatments using targeted agents (i.e. vascular endothelial growth factor inhibitors) might increase the number of patients for whom resection of clinically isolated LRs would be a reasonable option to consolidate their remission.

In conclusion, LR after radical nephrectomy is rare (2.9%) and has a poor prognosis. The presence of synchronous metastasis and nonsurgical therapy are associated with low survival rates. However, if there is no metastatic disease, complete surgical resection of isolated LRs might prolong disease-specific survival. The present data and similar data from other studies suggest that surgery for LRs should be restricted to patients with an isolated LR. All others should be offered systemic therapy.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES