BJU International

Preoperative metastatic status, level of thrombus and body mass index predict overall survival in patients undergoing nephrectomy and inferior vena cava thrombectomy

Authors


Philippe E. Spiess, Department of Genitourinary Oncology, Moffitt Cancer Center, 12902 Magnolia Drive, Office 12538, Tampa, FL 33612, USA. e-mail: philippe.spiess@moffitt.org

Abstract

Study Type – Prognosis (case series)

Level of Evidence 4

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

Little is known about the prognostic impact of body mass index (BMI) and obesity on patients with locally advanced kidney cancer. Previous studies suggest that clinical/pathological stage, the proximal extent of the tumour thrombus, direct vascular wall invasion, and preoperative performance status may all constitute important prognostic factors within this patient population.

The present study shows that a patient's metastatic status, higher level of tumour thrombus, and lower BMI all constitute adverse predictors of overall survival in patients who have RCC with inferior vena cava tumour thrombus.

OBJECTIVE

  • • To determine which clinical variables, including body mass index (BMI), predict overall survival (OS) after nephrectomy with inferior vena cava (IVC) thrombectomy for renal cell carcinoma (RCC) with tumour thrombus.

PATIENTS AND METHODS

  • • After institutional review board approval, a retrospective analysis of all patients (N= 100) undergoing nephrectomy and IVC thrombectomy for RCC from 1989 to 2010 were reviewed. One patient was excluded owing to missing clinical information leaving 99 patients in the study cohort.
  • • Patients were placed into one of two subgroups, based on their preoperative BMI (BMI ≤30 kg/m2 or BMI >30 kg/m2).
  • • Complications, blood loss, level of tumour thrombus, side of tumour and follow-up data were tabulated.

RESULTS

  • • Fifty-six patients had a BMI ≤30 kg/m2 and 43 patients had a BMI >30 kg/m2. Intraoperative complications occurred in 14% of those with BMI >30 kg/m2 and 5.4% of those with a BMI ≤30 kg/m2 (P= 0.171).
  • • On multivariate analysis, a higher thrombus level (III/IV vs I/II) and the presence of metastatic disease at time of diagnosis was associated with a worse OS (P= 0.041 and P < 0.001, respectively).
  • • The subgroup with a higher preoperative BMI had a significantly better OS (hazard ratio 0.42; 95% confidence interval 0.22–0.80, P= 0.009).
  • • Similarly, our Kaplan–Meier survival analysis showed an improved OS in the patient cohort with a BMI >30 kg/m2 (P= 0.016).

CONCLUSION

  • • Important predictors of outcome in patients undergoing nephrectomy with IVC thrombectomy for RCC with tumour thrombus include preoperative BMI, level of IVC tumour thrombus, and metastatic status at time of surgery.
Abbreviations
BMI

body mass index

OS

overall survival

IVC

inferior vena cava

ECOG

Eastern Cooperative Oncology Group

INTRODUCTION

Renal cell carcinoma with inferior vena cava (IVC) tumour thrombus can represent a surgical challenge, even at highly specialized tertiary care referral centres [1]. In addition, high perioperative morbidity has traditionally characterized such surgical procedures, particularly as the level of tumour thrombus extends into the intrahepatic and supradiaphragmatic portions of the IVC [2]. The incidence of morbid obesity continues to rise within the USA, with an estimated 35–40% of adults currently deemed to be obese [3]. In recent years, there has been increasing interest in deciphering the possible association of obesity with the incidence and outcome of genitourinary malignancies [4]. Previous studies have suggested that a higher body mass index (BMI), a surrogate marker of obesity, is associated with a more favourable clinical outcome in patients with localized RCC [5,6]. It has been proposed that an upregulation of leptin and downregulation of adinopectin pathways in obese patients may affect HIF1/2α levels and angiogenesis and thereby the cancer-specific outcomes of RCC [7].

It is known that 4–10% of patients with RCC present with venous tumour thombus for which the prognosis has been predominantly determined by preoperative performance status, lymph node or distant metastatic disease, macroscopic necrosis within the primary renal neoplasm and tumour grade [8–12]. The prognostic role of obesity in such patients has never been studied and remains to be determined. The aim of the present study was to determine the impact of obesity and other perioperative variables on the treatment outcomes of patients undergoing surgical resection for RCC with IVC tumour thrombus.

PATIENTS AND METHODS

Before conducting the present study, a retrospective chart review protocol was approved by our institutional review board to identify patients undergoing surgical management for RCC at our tertiary care referral centre. From November 1989 to July 2010, 100 consecutive patients underwent a radical nephrectomy and IVC thrombectomy for RCC. All patients underwent a complete metastatic evaluation within 4–6 weeks of their surgical resection, including a history, a physical examination, and serological studies including serum creatinine, complete blood count, calcium assessment, as well as liver function studies. In addition, all patients underwent preoperative radiographic imaging studies of the chest (x-ray, CT or MRI) and abdomen (CT or MRI with i.v. contrast). Additional imaging was performed at the discretion of the treating urologist based on the suspicion of occult metastatic disease at other sites. Of the 100 patients undergoing nephrectomy and IVC thrombectomy, 32 patients had metastatic disease at the time of diagnosis. Most of these patients were seen by both a urologist and genitourinary medical oncologist before surgery to assess their suitability for upfront surgical resection, with all patients having an Eastern Cooperative Oncology Group (ECOG) performance status <2 and deemed to be appropriate surgical candidates. None of the patients included in the present study had received preoperative systemic therapy (immunotherapy, targeted therapy and/or chemotherapy). Twenty five of the 32 (78.1%) with known metastatic disease received postoperative salvage systemic therapy (seven did not owing to poor performance status, their personal wishes or rapid disease progression prompting palliative/supportive measures). Obesity was assessed as a clinical variable, using BMI as its surrogate endpoint, with patients who had a calculated preoperative BMI >30 kg/m2 considered to be obese, according to a previously published definition [2]. One patient had no information available pertaining to his preoperative BMI and was therefore excluded from our data analysis. Our study population consisted of the remaining 99 patients who underwent nephrectomy and IVC thrombectomy for RCC. After surgical resection, patients were followed routinely every 3–6 months, with history, physical examination, serological testing and radiographic imaging of the chest (chest x-ray, CT or MRI), and abdomen (CT or MRI with i.v. contrast provided the creatinine clearance was adequate).

The level of the IVC venous tumour thrombus was reported in all cases according to the Mayo staging system in which a level 1 thrombus extends into the IVC no more than 2 cm above the origin at the renal vein ostium, level 2 extends into the IVC more than 2 cm above the renal ostium, but not within the intrahepatic portion of the IVC, level 3 extends into the intrahepatic IVC but remains below the diaphragm, and level 4 extends into the IVC above the diaphragm [13].

STATISTICAL ANALYSIS

Patient demographics and clinical characteristics were summarized using descriptive statistics stratifying patients into two subgroups based on their preoperative BMI (<30 kg/m2 and >30 kg/m2). Differences between these two BMI subgroups were compared using the t-test for continuous factors and the Fisher's exact test for categorical variables [14]. Potential prognostic factors of overall survival (OS) among patients undergoing nephrectomy and IVC thrombectomy were evaluated using Cox regression models. We observed that there were very few patients (n= 4) with follow-up data beyond 8 years, so we truncated our survival analysis at this time point. Both univariate and multivariate models of potential prognostic factors of OS were conducted, with those potential prognostic factors with a P value of <0.1 in the univariate model included in the multivariate Cox model. OS was defined as the interval between the dates of diagnosis with RCC until death. The survival curves stratified by the BMI and IVC tumour thrombus level subgroups were generated using the Kaplan–Meier method and were statistically compared with one another using the log-rank test.

RESULTS

The patient characteristics of our study population (N= 99) are shown in Table 1. The primary tumour histology was clear cell in 56 (75.7%), papillary in seven (9.5%), chromophobe in two (2.0%), and mixed histology in nine (12.2%) patients, with histology reporting not available in 25 patients. The preoperative ECOG performance status of patients was 0 in 65 (65.7%) and 1 in 34 (34.3%) patients. Among the 32 patients with known metastatic RCC before surgery, 22 fell into the favourable, eight into the intermediate, and two into the poor Motzer risk groups [15]. Of our study population, 56 patients had a BMI <30 kg/m2 and 43 had a BMI >30 kg/m2. The mean (sd) BMI for each group was 25.5 (3.1) and 35.0 (4.7), respectively (P < 0.001). The median (sd) age of patients at time of diagnosis was 64.0 (10.7) years. IVC tumour thrombus found in patients at time of diagnosis was level 1 in 23, level 2 in 35, level 3 in 22 and level 4 in 19 patients. Of the patients undergoing nephrectomy and IVC thrombectomy, intraoperative complications developed in nine patients and postoperative complications in 21. At a median (range) follow-up of 3.7 (0.2–8) years after surgery, 27 patients were alive without disease, 25 alive with disease, 41 had died from disease, four had died from other causes, and two had died from unknown causes.

Table 1. Overall summary of patient and clinical characteristics (N= 99)
Variable 
Mean (sd) age at diagnosis, years*64.0 (10.7)
Level of tumour thrombus, n (%) 
 I23 (23.2)
 II35 (35.0)
 III22 (22.0)
 IV19 (19.0)
Metastatic disease at Dx, n (%)32 (32.3)
Median (sd) blood loss, mL2754.3 (2319.6)
Intraoperative complications, n (%)9 (9.0)
Postoperative complications, n (%)21 (36.2)
Disease status, n (%) 
 Alive without disease27 (27.3)
 Alive with disease25 (25.3)
 Dead from disease41 (41.4)
 Dead from other causes4 (4.0)
 Dead from unknown causes2 (2.0)

Table 2 shows a comparison of the patient and clinical characteristics of the two BMI subgroups, with no significant differences between the two groups found in any of the clinical and pathological variables. The last follow-up disease status was significantly different (P= 0.023) in the two BMI subgroups, with 51.8% and 27.9% of patients dead from disease at last follow-up in the subgroups of patients with a BMI ≤30 kg/m2 and >30 kg/m2, respectively. The median (sd) follow-up time in the BMI ≤30 kg/m2 group was 44.3 (44.7) months and for the BMI >30 kg/m2 group it was 38.4 (33.5) months, with the difference in follow-up duration not significantly different (P= 0.135) for those who were alive at the end of the study.

Table 2. Comparison of patient and clinical characteristics in each BMI subgroup
 Subgroup P
BMI ≤30 kg/m2,n= 56BMI >30 kg/m2,n= 43
Mean (sd) age at diagnosis, years*65.3 (10.5)62.2 (10.9)0.204
Level of tumour thrombus, n (%)   
 I14 (25.0)9 (20.9)0.470
 II16 (28.6)19 (44.2) 
 III14 (25.0)8 (18.6) 
 IV12 (21.4)7 (16.3) 
Metastatic disease at diagnosis, n (%)17 (30.4)15 (35.7)0.665
Median (sd) blood loss, mL2487.5 (2407.7)3098.6 (2214.4)0.260
Intraoperative complications, n (%)3 (5.4)6 (14.0)0.171
Postoperative complications, n (%)12 (35.3)9 (39.1)0.787
Disease status, n (%)   
 Alive without disease13 (23.2)14 (32.6) 0.023
 Alive with disease9 (16.1)16 (37.2) 
 Dead from disease29 (51.8)12 (27.9) 
 Dead from other causes3 (5.4)1 (2.3) 
 Dead from unknown causes2 (3.6)0 (0) 

A Cox univariate and multivariate analysis of potential predictors of OS was conducted as shown in Table 3. On multivariate analysis, a higher level of IVC thrombus (III/IV vs I/II) and the presence of metastatic disease at time of diagnosis were associated with a higher risk of death (P= 0.041 and <0.001, respectively). In addition, patients with a higher preoperative BMI (>30 vs ≤30 kg/m2) had a better OS on multivariate analysis (P= 0.009).

Table 3. Cox univariate and multivariate analysis of prognostic factors associated with overall survival
VariablesUnivariate modelMultivariable model, n= 98
Crude HR (95% CI) P Adjusted HR (95% CI) P
Age at diagnosis1.01 (0.99, 1.04)0.357 
Level of thrombus   
 III/IV vs I/II1.73 (0.98, 3.04)0.0591.84 (1.03, 3.30) 0.041
Metastatic status2.32 (1.32, 4.10) 0.004 2.97 (1.65, 5.36) <0.001
Blood loss, per 1000 mL1.18 (1.04, 1.34) 0.011  
Intraoperative complication1.46 (0.62, 3.45)0.386 
Postoperative complication2.35 (1.19, 4.64) 0.014  
BMI, >30 vs ≤30 kg/m20.46 (0.25, 0.88) 0.019 0.42 (0.22, 0.80) 0.009

As shown in the Kaplan–Meier plot in Fig. 1, patients with a lower IVC tumour thrombus level (I/II) tended to have a better OS (P= 0.056) than those with a higher level of thrombus (level III/IV), with their median survival times being 6.6 years (95% CI 4.1– not available) and 1.4 years (95% CI 0.9–4.8), respectively. As for the Kaplan–Meier plot of OS stratified by preoperative BMI as shown in Fig. 2, patients with a BMI >30 kg/m2 had a better OS than those with a BMI ≤30 kg/m2 (P= 0.016). The median survival time of the subgroup having a BMI ≤30 kg/m2 was 1.8 years (95% CI = 1.3–5.5 years) and that of the >30 kg/m2 subgroup was not reached.

Figure 1.

Kaplan–Meier survival curve of OS stratified by level of IVC tumour thrombus. The numbers of patients at risk of death at the various time points in both subgroups are included at the bottom of the graph.

Figure 2.

Kaplan–Meier survival curve of OS stratified by the preoperative BMI subgroups. The numbers of patients at risk of death at the various time points in both subgroups are included at the bottom of the graph.

DISCUSSION

The present study shows that the presence of metastatic disease at time of surgery, higher level (III/IV) of IVC tumour thrombus, and lower preoperative BMI (≤30 kg/m2) are all adverse predictors of OS in patients undergoing nephrectomy and IVC thrombectomy for RCC. In an attempt to ensure that the adverse prognosis of having a BMI ≤30 kg/m2 was not attributable to some patients being cachectic before surgical resection, we ensured the data were not skewed by poorly nourished patients being present among the subgroup with a BMI ≤30 kg/m2. In fact, only one patient had a preoperative BMI <18 kg/m2 (BMI 17.4 kg/m2), which is considered the reference value for being considered malnourished/underweight, so this did not significantly impact our results. Similarly, all of our patients had an ECOG preoperative performance status of <2, indicating all patients were suitable surgical candidates. Based on our results, we reviewed the scientific literature to determine what was known about the clinical association between obesity and RCC. In a previous study by Parker et al. [5] the association between BMI and localized RCC was assessed only among patients who had clear-cell histology. The authors were able to show that overweight and obese patients were more likely to exhibit less aggressive final tumour pathology and improved cancer-specific outcomes vs normal weight patients. In a study by Haferkamp et al. [6], the authors were able to corroborate this among patients with localized RCC; overweight patients had a strong tendency to less aggressive disease. Furthermore, obesity has recently been shown to be a favourable prognostic factor in patients with metastatic RCC treated with novel systemic agents such as vascular-endothelial-growth factor-targeted agents [16]. The present study, however, is the first to evaluate the association between preoperative obesity and survival in patients with locally advanced RCC (clinical stage T3b-c) and raises the question of why there would be a more favourable outcome with the surgical management of localized and locally advanced disease in patients with preoperative obesity. The limited pathophysiological research on this topic would suggest that the presence of obesity affects the leptin and adinopectin pathways which are thought to be implicated in RCC pathogenesis via the HIF1/2α pathways [5]. The novel clinical findings of the present study need to be investigated in translational studies which may help further characterize the pathophysiological basis of RCC and identify potential targets for future therapeutic strategies.

The present study highlights the fact that important predictors of OS among patients undergoing nephrectomy and IVC thrombectomy include the presence of metastatic disease at time of diagnosis and the level of the IVC tumour thrombus. In the present study, patients with metastatic disease at time of diagnosis with RCC and an IVC thrombus had a hazard ratio of 2.97 and were nearly three times more likely to die than those without metastatic disease. Our findings corroborate the results of Tanaka et al. [17], who reported that patients with RCC with both IVC tumour thrombus and nodal metastasis had a poor prognosis (OS at 5 years of only 32.9%). This clearly has important clinical implications and we would propose that those with a significant metastatic tumour burden at time of diagnosis should be counselled about their expected survival in light of a potentially morbid surgical procedure.

In addition, a higher level of IVC tumour thrombus (level III/IV) was associated with a poorer OS on multivariate analysis. There has been some controversy in recent years on the prognostic importance of the level of IVC tumour thrombus on OS, with a recent multicentre study of 1192 patients by Wagner et al. [9] concluding that the level of tumour thrombus within the IVC does not significantly affect OS within this patient cohort. However, in a subsequent study based on a pooled analysis of 11 international centres, the level of IVC tumour thrombus was shown to be an independent predictor of OS on multivariate analysis [18]. The present study corroborates the prognostic importance of the level of IVC tumour thrombus within this patient cohort.

There are limitations to the present study that must be acknowledged. Firstly, this was a retrospective study with an inherent selection bias that cannot be overcome. Secondly, the pathophysiological basis for the association between obesity and survival among patients with localized and locally advanced RCC will need to better characterized from a pathophysiological standpoint for our work to have not only epidemiological but therapeutic implications.

In conclusion, the present study shows that a lower preoperative BMI (≤30 kg/m2), the presence of metastatic disease at the time of diagnosis, and higher level of IVC tumour thrombus (level III/IV) predict poorer OS among patients undergoing nephrectomy and IVC thrombectomy for RCC. These important predictors of survival must be taken into account when educating and tailoring treatment recommendations within this patient cohort.

CONFLICT OF INTEREST

None declared.