Influence of nutritional deficiency on prognosis of renal cell carcinoma (RCC)

Authors


Correspondence: Hyeon H. Kim, Department of Urology, Seoul National University Hospital, 28, Yongon-dong, Jongno-gu, Seoul 110-744, Korea.

e-mail: hhkim@snu.ac.kr

Abstract

Objective

  • To evaluate the prognosis of patients with renal cell carcinoma (RCC) by nutritional status defined by body mass index (BMI), serum albumin and cholesterol.

Patients and Methods

  • This study retrospectively enrolled 1437 patients who underwent radical nephrectomy (932) or partial nephrectomy (505) for RCC.
  • We assigned nutritional status according to the presence of none or one nutritional risk factor (control group) and two or all three of the following nutritional risk factors (nutritional deficiency group).
  • The nutritional factors and thresholds were preoperative albumin level (<3.5 g/dL), preoperative cholesterol level (<220 mg/dL), and preoperative BMI (<23 kg/m2)

Results

  • The patients' mean (sd) age was 55.23 (12.41) years and BMI was 24.36 (3.17) kg/m2. The mean (sd) serum cholesterol level was 180.07 (38.24) mg/dL, and the albumin level was 4.2 (0.45) g/dL.
  • In all, 141 (9.8%) patients had none of the nutritional deficiency criteria, 802 (55.8%) had one, 429 (29.9%) had two, and 65 (4.5%) had all three.
  • Clinicopathological variables, i.e. female gender, high tumour stage, positive lymph node metastasis, positive distant metastasis, high nuclear grade and non-clear cell type histopathology were associated with the nutritional deficiency group.
  • In multivariate Cox analysis, nutritional deficiency was an independent predictor for RCC recurrence (hazard ratio [HR] 1.39, 95% confidence interval [CI] 1.05–1.83, P = 0.020) and RCC-related mortality (HR 2.06, 95% CI 1.39–3.03, P < 0.001).

Conclusion

  • Nutritional deficiency defined by BMI, serum albumin and cholesterol is an important factor that predicts postoperative prognosis of patients with RCC who have undergone radical or partial nephrectomy.

Introduction

Tumour cells in RCC are well known for controlling metabolic pathways. As we learn even more about the genetic molecular basis of RCC, we have found that mutations of metabolic genes related to oxygen, iron, energy or nutrient sensing affect development and aggressiveness of RCC [1]. This means that RCC is a metabolic disease. In this respect, nutritional status related to cellular metabolism can be an important prognostic factor of RCC.

Nutritional status is an important risk factor related to the prevalence of postoperative complications and disease-related mortality not only in gastrointestinal but also in urological malignancies [2-6]. However, the results of numerous criteria for assessing nutritional deficiency, i.e. the nutritional risk index (NRI), nutritional risk score (NRS) and subjective global assessment (SGA), are potentially subjective, as they are mainly based on patient's memory. Therefore, we need simple and objective criteria that reveal the patient's nutritional status.

Pretreatment serum albumin is a very useful prognostic factor in various malignancies including RCC, and body mass index (BMI) is related to the postoperative prognosis of RCC [7-9]. We recently showed that overweight and obese patients with RCC have more favourable pathological features and a better prognosis [10]. As cholesterol level is associated with mortality in medical and advanced stage cancer patients, we hypothesised that cholesterol level would be a valuable factor in predicting the survival of patients with RCC [11, 12]. Although each of BMI, albumin, and cholesterol has a valid value for the evaluation of nutritional status, if we create a new criterion using each of those parameters as a single risk factor, we will be able to evaluate nutritional status more accurately.

In the present study, we defined nutritional deficiency using BMI, serum albumin and cholesterol levels that are easily obtainable and reproducible in the clinical setting. The prognostic value of nutritional deficiency was evaluated in patients with RCC who had undergone curative surgery.

Patients and Methods

We performed a retrospective study of 1437 patients who underwent radical nephrectomy (932) or partial nephrectomy (505) for RCC at our hospital from January 1999 to December 2009. Institutional Review Board approval was obtained for the retrospective analysis of this patient population.

Preoperatively we evaluated the patient's gender, age, underlying disease, American Society of Anesthesiologists (ASA) score and preoperative laboratory results, i.e. complete blood cell count, serum chemistry (calcium, albumin, creatinine level), c-reactive protein, and erythrocyte sedimentation rate. Tumour size and location was identified by CT. After surgery, a pathologist evaluated the surgical specimens according to the 2010 American Joint Committee on cancer guidelines and Fuhrman grading system. Histological subtyping was classified by the 2004 WHO classification. Survival and disease progression data were gathered by medical record review, by calling patients families or by reviewing death certificates. Follow-up was recorded from the operation date to the date of death or last follow-up visit.

BMI was classified as a four-level categorical variable according to WHO Asian-pacific area and Korean Society for the Study of Obesity Scale [13]. Hypoalbuminaemia is a prognostic factor with known effects on mortality and morbidity of RCC, so we divided albumin level in to two categories with a threshold of 3.5 g/dL. Hypercholesterolaemia (>220 mg/dL) is a risk factor for metabolic syndrome, and hypocholesterolaemia (<150 mg/dL) is a predictor of high mortality in advanced disease [11, 12]. In the present study, the mean cholesterol level was 180.07 mg/dL, so we categorised cholesterol level into four groups (Table 1).

Table 1. Risk factors of nutritional deficiency
VariableStrataNo. of patients (%)
Albumin, g/dL<3.5118 (8.2)
≥3.51319 (91.8)
BMI, kg/m2<18.525 (1.7)
18.5 to 22461 (32.1)
23 to 24393 (27.3)
≥25558 (38.8)
Total cholesterol, mg/dL<150288 (20.0)
150 to 179419 (29.1)
180 to 219544 (37.9)
≥220186 (12.9)

To confirm nutritional deficiency, we divided our patients into two groups according to the presence of none or one of the following nutritional risk factors (control group) and two or all three of the following factors (nutritional deficiency group): preoperative albumin level (<3.5 g/dL), preoperative cholesterol level (<220 mg/dL), preoperative BMI (<23 kg/m2) (Table 2).

Table 2. Definition of nutritional deficiency group
 Number of risk factorsNo. of patients (%)No. of patients (%)
Control group0141 (9.8)943 (65.6)
1802 (55.8)
Nutritional deficiency group2429 (29.9)494 (34.4)
365 (4.5)

To measure the relationship between nutritional deficiency and clinicopathological variables, chi-square and Mann–Whitney tests were used. The Kaplan–Meier method and the log-rank test were used for univariate survival analysis. For multivariate analysis, Cox proportional hazard regression was used for cancer-specific survival (CSS) and progression-free survival (PFS). Each hazard ratio (HR) was given, including the 95% CI. In all tests, a P < 0.05 was considered to indicate statistical significance.

Results

The mean (sd) age was 55.23 (12.41) years and the gender split was 1028 men (71.5%) and 409 women (28.5%). The mean (sd) BMI was 24.36 (3.17) kg/m2, mean serum cholesterol was 180.07 (38.24) mg/dL and mean albumin level, 4.2 (0.45) g/dL. On final pathology there were 1151 (80.1%) clear cell, 130 (9.0%) chromophobe, 45 (3.1%) papillary I, 48 (3.3%) papillary II, and six (0.4%) collecting duct carcinomas (Table 3). In all, 232 (16.1%) patients relapsed and 115 (8%) patients had died from RCC. The median (interquartile range) follow-up was 36.0 (16–62) months.

Table 3. Clinicopathological variables
VariableValue
Sex M/F, n (%)1028 (71.5)/409 (28.5)
Mean (sd): 
Age, years55.23 (12.41)
BMI, kg/m224.36 (3.17)
Serum albumin, g/dL4.2 (0.45)
Serum cholesterol, mg/dL180.07 (38.24)
Pathology, n (%): 
Clear cell1151 (80.1)
Chromophobe130 (9.0)
Papillary I/II45 (3.1)/48 (3.3)
Collecting duct6 (0.4)

There were 141 (9.8%) patients who had no nutritional deficiency factors, 802 (55.8%) had one, 429 (29.9%) had two, and 65 (4.5%) had all three (Table 2). Clinicopathological variables, i.e. female gender, high tumour stage, positive lymph node metastasis, positive distant metastasis, high nuclear grade and non-clear cell type histopathology were associated with the nutritional deficiency group (Table 4).

Table 4. The nutritional deficiency group was associated with the female gender, TNM stage, high nuclear grade and non-clear cell type histopathology
  No. of patients (%)Number of risk factors, n (%)P
0–1 (control)2–3 (ND)
  1. chi-square and Mann-Whitney test; ND, nutritional deficiency group; Nx, nephrectomy.
SexMale1028 (71.5)699 (68.0)329 (32.0)0.003
Female409 (28.5)244 (59.7)165 (40.3)
Age, years<56712 (49.5)470 (66.0)242 (34.0)0.781
≥56725 (50.5)473 (65.2)252 (34.8)
ASA score1660 (45.9)430 (65.2)230 (34.8)0.799
2692 (48.2)467 (67.5)225 (32.5)
3/481/4 (5.9)46 (54.1)39 (45.9)
T stage11059 (73.7)728 (68.7)331 (31.3)<0.001
2129 (9.0)85 (65.9)44 (34.1)
3219 (15.2)119 (54.3)100 (45.7)
430 (2.1)11 (36.7)19 (63.3)
LN metastasis01369 (95.7)907 (66.3)462 (33.7)0.013
Nx62 (4.3)31 (50.0)31 (50.0)
Distant metastasis01315 (91.8)887 (67.5)428 (32.5)<0.001
Mx117 (8.2)52 (44.4)65 (55.6)
Furman nuclear grade143 (3.0)31 (72.1)12 (27.9)0.003
2703 (49.6)481 (68.4)222 (31.6)
3570 (40.2)365 (64.0)205 (36.0)
4101 (7.1)53 (52.5)48 (47.5)
Histopathologyclear cell1151 (80.2)773 (67.2)378 (32.8)0.015
non-clear cell285 (19.8)169 (59.3)116 (40.7)
OperationRadical Nx.932 (64.9)597 (64.1)335 (35.9)0.092
Partial Nx.505 (35.1)346 (68.5)159 (31.5)

There was a significant difference in the 5-year PFS between patients with a BMI of >23 kg/m2 and those with a BMI of ≤23 kg/m2 (log-rank test, 82.8% vs 76.2%, P < 0.001, data not shown). High preoperative albumin and cholesterol levels were also associated with better 5-year PFS (log-rank test, albumin; 82.6% vs 56.1%, P < 0.001, cholesterol; 86.8% vs 73.9%, P < 0.001, data not shown). Kaplan–Meier analysis showed that the 5-year CSS was 91.7% for patients with a BMI of >23 kg/m2 and 84.4% for patients with a BMI of ≤23 kg/m2 (log-rank test, P < 0.001, data not shown). Patients with an albumin level of >3.5 g/L and cholesterol level of >220 mg/L also had a significantly higher 5-year CSS (log-rank test, albumin; 91.2% vs 65.7%, P < 0.001, cholesterol; 95.4% vs 88.3%, P = 0.045, data not shown). Compared with the nutritional deficiency group, the control group had favourable outcomes in 5-year PFS (84.1% vs 73.6%, P < 0.001, Fig. 1A) and in 5-year CSS (92.9% vs 82.0%, P < 0.001, Fig. 1B). In multivariate Cox analysis, nutritional deficiency was an independent predictor for recurrence of RCC (HR 1.39, 95% CI 1.05–1.83, P = 0.020; Table 5) and RCC-related mortality (HR 2.06, 95% CI 1.39–3.03, P < 0.001; Table 5).

Figure 1.

Compared with the nutritional deficiency group, the control group had favourable outcomes for 5-year PFS (A) and 5-year CSS (B).

Table 5. Cox analysis, nutritional deficiency was an independent predictor for recurrence of RCC and RCC-related mortality
 PFSCSS
HR (95% CI)PHR (95% CI)P
Sex0.93 (0.68–1.26)0.6181.15 (0.75–1.78)0.519
Age1.21 (0.91–1.61)0.1851.07 (0.70–1.63)0.754
ASA score    
1Reference Reference 
21.21 (0.89–1.63)0.2271.11 (0.70–1.74)0.664
3/41.52 (0.96–2.41)0.0761.43 (0.74–2.74)0.284
T stage    
1Reference Reference 
23.82 (2.53–5.76)<0.0014.94 (2.53–9.66)<0.001
36.43 (4.50–9.19)<0.0017.48 (4.09–13.69)<0.001
44.94 (2.75–8.87)<0.0018.18 (3.80–17.62)<0.001
N stage0.96 (0.63–1.47)0.8611.39 (0.83–2.31)0.207
M stage7.25 (4.97–10.58)<0.0017.22 (4.45–11.70)<0.001
Nuclear grade2.17 (1.55–3.03)<0.0013.42 (1.94–6.04)<0.001
Histopathology0.92 (0.64–1.31)0.6360.93 (0.55–1.58)0.795
Nutritional status1.39 (1.05–1.83)0.0202.06 (1.39–3.03)<0.001

Discussion

Advanced urological surgical skills have made it possible for older patients, those in poor general condition, and those with co-morbidities to have better outcomes after curative surgical treatment. Thus, it is necessary to evaluate factors like nutritional status that can help to predict the postoperative course in operable patients with RCC. However, there is not a simple standardised method to assess nutritional status of patients with RCC.

NRS, NRI, and SGA are widely used for the assessment of nutritional deficiency. The NRS consists of BMI, weight loss, poor appetite, ability to eat, and severity of disease. The NRI consists of recent weight loss and serum albumin concentration. The SGA has similar criteria, i.e. weight change, dietary intake change, gastrointestinal symptom, functional capacity. The limitation of these criteria is that they have to rely on patient's subjective memory. Serum albumin, BMI, and total cholesterol level can be good nutritional deficiency risk factors in patients with RCC, as they are easily obtainable and objective. Moreover, these three factors are already known to influence morbidity and mortality in other diseases [7-9, 11, 12, 14].

Serum albumin is the most abundant protein in human plasma and reflects visceral protein function. Serum albumin level provides useful prognostic value in cancer [14-16]. Tumour cells release interleukin 6, pro-inflammatory cytokines, and growth factors that have a profound catabolic effects on patient metabolism [17], and the permeability of vessels is increased TNF, so that serum albumin is easily passed to the third space through the microvasculature. As a result, serum albumin levels decrease significantly as cancer progresses [15, 16]. Serum albumin is not a perfect prognostic factor because of a long half-life and because synthesis is suppressed by malnutrition and inflammation [15, 18]. However, serum albumin level is a valuable prognostic factor for patients with RCC because it is readily available, inexpensive and reproducible [19].

Some studies have reported that being overweight or obese were significant risk factors for RCC [20-24]. Hypertension, cholesterol metabolism abnormalities, immune malfunction, elevation of humoral growth factors in adipose tissue, higher oestrogen levels, and higher insulin levels lead to poor prognosis [25]. Paradoxically, recent studies have reported thatpatients with RCC with high BMIs have a more favourable prognosis than those with a normal BMI [7-9, 26-28]. In the present study, being overweight or obese preoperatively (BMI >23 kg/m2) was associated with better PFS (82.8% vs 76.2%, P < 0.001) and CSS (91.7% vs 84.4%, P < 0.001). The increased levels of IGF-1, leptin, and pre-albumin in obese patients might be associated with better prognosis [29], but the mechanism is unclear.

RCC is involved in metabolism through aerobic glycolysis and lipid metabolism [30]. Particularly, RCC cells contain a lot of lipid, as they store lipids after converting free cholesterol to esterified cholesterol using acetyl-CoA cholesterol acetyl transferase (ACAT) [31] at a rate 5.7-fold higher than normal kidney cells. Moreover, unlike normal cells, RCC cells can produce fatty acids de novo with the help of fatty acid synthase (FASN) [32, 33]. Thus, if cholesterol levels of patients are affected by the ACAT or FASN activity in RCC, it might be possible to predict disease-free survival in RCC just as in other advanced disease [11, 12]. In the present study, we were able to collect serum cholesterol levels in all 1437 patients. Cholesterol levels showed a normal distribution, with a mean (sd) of 180.1 (38.24) mg/dL. We divided patients into two groups, lower and higher cholesterol level groups by applying three different threshold values: 150 mg/dL, 180 mg/dL, and 220 mg/dL. With any threshold value, a higher cholesterol level was associated better PFS and CSS.

Nutritional deficiency of a patient is influenced by various factors and it is not easy to fully understand how nutritional deficiency affects a patient. However, the importance of the result of the present study is that the data helps define the ideal patient population for modification of nutritional status. In the present study, nutritional deficiency was associated with progression of RCC and RCC-related mortality. Therefore, an alternative means of modification of their nutritional status, which may decrease mortality, might be considered earlier for this population of patients in whom the disease is thought to be biologically aggressive. However, further prospective studies are needed to confirm the present retrospective data for prospectively supporting our results and for defining the proper thresholds of the nutritional deficiency variables.

The main limitation of the present study is that it was retrospective and conducted in a single centre. However, 1437 patients with RCC were enrolled and nearly complete laboratory data by medical record review was available. Many factors that could influence the results were well controlled for in the present analysis. It is unlikely that the apparent survival differences associated with nutritional status were an artifact of an imbalance of prognostic factors. A second weakness was that we did not consider the patient's weight loss and patient's presenting symptoms. However, as we mentioned previously, these factors rely on patient's subjective memory. Also, in the era of small renal mass, the importance of weight loss and subjective symptoms for RCC might be reduced in contrast to previous eras. The present data included 505 partial nephrectomies and nearly all were diagnosed incidentally during medical examination for unrelated conditions. The results of the present study may be helpful for the evaluation of the nutritional deficiency of patients.

In conclusion, nutritional deficiency defined by BMI, serum albumin and cholesterol levels, is an important factor that predicts postoperative prognosis of patients with RCC who have undergone radical or partial nephrectomy.

Acknowledgements

There is no financial support or incentive for this manuscript.

Conflict of Interest

None declared.

Abbreviations
ACAT

acetyl-CoA cholesterol acetyl transferase

ASA

American Society of Anesthesiologists

CSS

cancer-specific survival

BMI

body mass index

FASN

fatty acid synthase

NRI

nutritional risk index

NRS

nutritional risk score

PFS

progression-free survival

SGA

subjective global assessment

Ancillary