SEARCH

SEARCH BY CITATION

Keywords:

  • renal cell cancer;
  • systemic inflammation;
  • biomarkers;
  • CRP;
  • WBC;
  • threshold;
  • predictive accuracy

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

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

White blood cell count and C-reactive protein are reliable prognostic RCC Biomarkers.Nevertheless, accepted cut-offs for risk stratifications are missing.

Therefore, both parameters were re-evaluated and multivariable analyses revealed an optimal CRP breakpoint at 0.25 mg/dL to be best to stratify patients at risk of cancer-specific mortality. However, this CRP-based prediction added no additional information compared to a clinico-pathological based model.

Objective

  • To re-evaluate the prognostic and predictive significance of the preoperative white blood cell (WBC) count and C-reactive protein (CRP) that independently predicts patient prognosis and to determine optimal threshold values for CRP.

Patients and Methods

  • From 1996 to 2005, 327 patients with surgery for clear cell renal cell carcinoma were retrospectively evaluated.
  • Cox proportional hazard models were used, adjusted for the effects of tumour stage, size, Fuhrman grade and Karnofsky index, to evaluate the prognostic significance of WBC count and CRP and to identify threshold values.
  • Identified thresholds were correlated with clinicopathological parameters and used to estimate cancer-specific survival.
  • To prove any additional predictive accuracy of the identified threshold it was compared with a clinicopathological base model using the Harrell concordance index (c-index).

Results

  • In univariable analyses WBC count was a significant prognostic marker at a concentration of 9.5/μL (hazard ratio [HR] 1.83) and 11.0/μL (HR 2.09) and supported CRP values of 0.25 mg/dL (HR 6.47, P < 0.001) and 0.5 mg/dL (HR 7.15, P < 0.001) as potential threshold values.
  • If adjusted by the multivariable models WBC count showed no clear breakpoint, but a CRP value of 0.25 mg/dL (HR 2.80, P = 0.027) proved to be optimal.
  • Reduced cancer-specific survival was proved for CRP 0.25 mg/dL (median 69.9 vs 92.3 months). Median follow-up was 57.5 months with 72 (22%) tumour-related deaths.
  • The final model built by the addition of CRP 0.25 mg/dL did not improve predictive accuracy (c-index 0.877) compared with the clinicopathological base model (c-index 0.881) which included TNM stage, grading and Karnofsky index.

Conclusions

  • Multivariable analyses revealed that an optimal breakpoint of CRP at a value of 0.25 mg/dL was best to stratify patients at risk of cancer-specific mortality, but CRP 0.25 mg/dL added no additional information in the prediction model.
  • Therefore we cannot recommend measuring CRP as the traditional parameters of TNM stage, grading and Karnofsky index are already of high predictive accuracy.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

RCC is the most frequent malignant tumour of the kidney with a rising incidence of 60 920 patients and 13 120 cancer-related deaths in the USA in 2011 [1]. The widespread use of ultrasound and axial imaging increased the ratio of localized tumours compared with advanced and metastatic RCC at initial diagnosis. Nevertheless cure is only available after complete surgical excision; and recurrence of RCC, even years after curative treatment, is not uncommon for this tumour entity, which occurs in one-third of patients. Then the prognosis of metastatic RCC is poor, despite recent advances in new targeted therapies to significantly increase response rates and to prolong progression-free and cancer-specific survival [2].

Prognostic stratification might allow the identification of patients at risk of tumour recurrence and therefore who could be recommended for a potential adjuvant therapy or individualized follow-up scheme, especially if preoperative prognostic models are used [3]. Current prognostic factors to predict the natural course of RCC include pathological characteristics like the TNM system, Fuhrman grading and tumour size, or additional patient-based factors like the Karnofsky status performance scale or distinctive laboratory values [4-6]. Prognostic models have already aimed to discriminate between favourable and unfavourable RCC phenotypes [7]. In addition, new molecular markers have been discussed to improve their predictive accuracy [8]. However, none of these molecular markers is recommended for routine practice due to high costs and limited availability or comparability [5].

Inflammation and cancer are closely linked together. As early as 1863, Rudolf Virchow noted leucocytes to be present in neoplastic tissue and suggested a link between inflammation and cancer development [9]. Later the observation of leucocyte invasion in stromal tissues of human malignancies was applied to serum white blood cell (WBC) count with enhanced tumour severity and reduced prognosis [10]. RCC is considered to be an immunological inflammatory tumour which can be targeted by immune therapy but which also manipulates the immune system to drive tumour escape mechanisms [11, 12]. The induction of peritumoural inflammation plays an important role in the host tumour defence response, resulting in induced synthesis of the acute phase reactant C-reactive protein (CRP) by hepatocytes after interleukin-6 stimulation from the tumour microenvironment [13]. Surrogate markers of inflammation are most commonly applied to monitor infectious disease, but they are more and more used to predict a progressive tumour disease in patients with cancer [14].

Non-specific inflammatory markers like WBC count or CRP have been integrated in daily clinical practice to measure inflammatory disorders. Knowing the above-mentioned background, non-specific inflammatory markers like WBC count or CRP are promising candidates to predict patients’ prognosis in oncological conditions.

The rise of these most commonly used surrogate markers of inflammation indicates the host reaction not only to infection but also to progressive tumour disease in nearly all patients with cancer. Here, CRP is highly associated with tumour progression and cancer survival [15, 16]. This close link between inflammation and immunological mechanisms has also been reported for RCC [11]. Consequently, increased CRP levels have been associated with advanced and progressive renal cancers [14] and preoperative CRP predicted metastasis and mortality after curative-intended nephrectomy [17]. CRP was also used to raise predictive accuracy for RCC-related death in a Japanese cohort of patients [18]. In particular, in the cytokine era the rise of immune parameters (high WBC count and intratumoural neutrophils) was correlated with poor outcome of immunotherapy in patients with metastatic RCC [19]. Consistently, a report from the Groupe Français d'Immunothérapie linked the highest relative risks of metastatic RCC progression to an elevated WBC count under cytokine regimens [20].

Despite numerous reports underlining the role of circulating inflammatory markers, it has not been questioned whether WBC count or CRP is the better one, nor have accepted threshold values been described. The aim of the present study was to re-evaluate the prognostic significance of preoperative WBC count and CRP that independently predict patient prognosis beyond other accepted factors and to determine optimal threshold values for patient stratification. Additionally we explored whether the obtained threshold value raises the predictive accuracy for RCC-related cancer mortality.

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

In this retrospective cohort, the clinical and pathological records of 327 consecutive patients who underwent partial or radical nephrectomy for clear cell RCC at our institution between 1993 and 2007 were reviewed. The study was approved by the ethics committee of the University of Tuebingen (078/2012BO2). In general patients were evaluated by axial imaging at the time of surgery and postoperatively every 3–4 months for the first year, semi-annually for the second and third years and annually by chest X-ray or thoracic CT, abdominal sonography, CT or MRI, and serum chemistry. Survival was calculated from the date of surgery. The survival endpoint was RCC-related death or the date of last follow-up. Data for patients who died from other than metastatic disease were censored at the time of death.

Clinical data were assessed for WBC count, CRP, age, gender, Karnofsky status performance scale, secondary malignancy, cause of death and follow-up time. Pathological data included TNM stage, Fuhrman nuclear grade [21] and tumour size. WBC count and CRP were measured 1–2 days before surgery. WBC count was calculated with an automated cell counter using 1.5 mL EDTA samples (reference 4.5–11.0/μL), while serum CRP levels (reference ≤0.5 mg/dL) were analysed by immunoturbidimetry using 2.0 mL lithium-heparin plasma samples.

Continuous normally distributed variables are given as mean value ± sd, continuous non-normally distributed variables as median values and interquartile ranges and categorical data as absolute and relative frequencies. Comparisons of categories of clinical and pathological data between the groups of patients were evaluated with Fisher's exact tests. Survival curves were estimated according to the Kaplan–Meier method and differences in survival were evaluated by the log-rank test. Survival rates at 5 years including 95% asymptotic confidence interval are given if appropriate.

To identify a potential threshold between different WBC count and CRP values, hazard ratios (HRs) for a range of possible breakpoints were estimated and tested in a univariable Cox proportional hazard regression model. The analyses were repeated with adjustment for other well known prognostic factors such as TNM stage, tumour size, Fuhrman grade and Karnofsky index in a multivariable Cox model.

Univariable predictive accuracy was determined for each variable and was defined as the ability to discriminate between patients who died from cancer. The predictive accuracy was evaluated using the Harrell concordance index (c-index) and is given as a percentage [22]. A clinicopathological (CP) base model was built, consisting of all variables that demonstrated a significant independent prognostic value in the multivariable Cox proportional hazard regression model. The predictive accuracy of the CP base model was compared on the addition of CRP. For internal validation of the predictive accuracy bootstrapping (200 resamples) was performed as suggested by Taylor et al. [23].

A two-sided probability value <0.05 (Statistical Package for Social Sciences Software, version 16.0, Chicago, IL, USA) was considered a significant effect, although this cannot be interpreted as confirmatory. The calculation of the Harrell c-index was performed using the statistical software package R.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

The mean age of the 327 patients was 63.5 (sd ± 12.0) years, and 67.0% (219 patients) were male. Median duration of follow-up was 57.5 (interquartile range 22.6–94.7) months. Of these, 128 patients (39.1%) died during follow-up, and tumour-related death occurred in 72 patients (22.0%). Patients were classified as pT1 (61.8%), pT2 (3.4%), pT3 (32.7%) and pT4 (2.1%). Lymph node involvement (pN1) was present in 19 patients (5.8%) and 43 patients (13.1%) had evidence of distant metastatic disease (cM1). Nuclear grading according to the Fuhrman classification was G1 in 61 (18.8%), G2 in 209 (64.3%) and G3/4 in the other patients (16.9%). Reduced performance status (Karnofsky <80%) was found in 94 patients (30.2%) (Table 1).

Table 1. Clinical and pathological characteristics of preoperative CRP values in 327 patients with clear cell RCC
Clinical and pathological characteristicsAllCRP
≤0.25 mg/dL>0.25 mg/dLP
Patients, n (%)327108 (33.0)219 (67.0) 
Age (years), mean ± sd63.5 ± 12.062.1 ± 13.464.1 ± 11.3 
Range17–9017–9028–89 
Sex, n (%)    
Male219 (67.0)74 (68.5)145 (66.2)0.709
Female108 (33.0)34 (31.5)74 (33.8) 
Histology parameters, n (%)    
T1/2213 (65.2)87 (80.6)126 (57.5)<0.001
T3/4114 (34.9)21 (19.4)93 (42.5) 
N0/X308 (94.2)108 (100.0)200 (91.3)0.002
N1/219 (5.8)019 (8.7) 
M0284 (86.9)104 (96.3)180 (82.2)<0.001
M143 (13.1)4 (3.7)39 (17.8) 
G1/2 (n = 325)270 (83.1)104 (97.2)166 (76.1)<0.001
G3/455 (16.9)3 (2.8)52 (23.9) 
Follow-up time (months)    
Median57.567.249.5 
Interquartile range22.6–94.749.7–101.415.8–92.9 
Follow-up: living patients67.071.565.3 
Interquartile range44.2–105.450.9–102.935.4–108.1 
Total deaths, n (%)128 (39.1)21 (6.4)107 (32.7)<0.001
Tumour-related deaths, n (%)72 (22.0)6 (1.8)66 (20.2)<0.001

Univariable analysis of different WBC values (9.5–11/μL) demonstrated a prognostic significant difference between the groups (≤9.5 vs >9.5, HR 1.83, 95% CI 1.1–2.9, P = 0.013; ≤10.0 vs >10.0, HR 1.24, 95% CI 0.72–2.1, P = 0.447; ≤11.0 vs >11.0, HR 2.09, 95% CI 1.2–3.7, P = 0.013; Table 2), but a clear breakpoint was not detectable. As shown in Table 2 CRP values of 0.15, 0.25 and 0.5 mg/dL led to the highest significant difference in prognosis between the respectively defined subgroups, with an HR of 7.09 (CRP 0.15 mg/dL), 6.47 (CRP 0.25 mg/dL) and 7.15 (CRP 0.5 mg/dL). To adjust for other prognostic factors beyond WBC count or CRP (see Table 3), a multivariable Cox model was applied and revealed a CRP level of 0.25 mg/dL to have the highest prognostic value with an HR of 2.8, P = 0.027, while no clear breakpoint for WBC count between the respective subgroups was detectable (Table 3). Patients with a CRP value of >0.25 mg/dL were associated with a significant positive correlation to a higher T stage, a higher nuclear grade, distant metastasis and positive regional lymph nodes (Table 1). The 5-year cancer-specific survival rate for patients with CRP ≤ 0.25 mg/dL was 92.3% (CI 89.1–95.5) and for those with CRP > 0.25 mg/dL it was 69.9% (CI 66.5–73.3) (Table 4; Fig. 1).

figure

Figure 1. Kaplan–Meier estimates of 327 patients with clear cell RCC categorized by a CRP value of 0.25 mg/dL, P < 0.001 (log-rank analysis); red line CRP ≤ 0.25 mg/dL; green line CRP > 0.25 mg/dL.

Download figure to PowerPoint

Table 2. Univariable and multivariable analysis of different inflammatory breakpoints of WBC and CRP in patients with clear cell RCC
CovariableCategoriesAll clear cell RCC, univariableAll clear cell RCC, multivariable
Hazardratio95% CIPHazardratio95% CIP
Leucocytes≤9.5 vs >9.51.831.1–2.90.0131.911.1–3.20.017
 ≤10.0 vs >10.01.240.72–2.10.4471.560.86–2.80.144
 ≤11.0 vs >11.02.091.2–3.70.0131.971.0–3.80.047
CRP≤0.05 vs >0.053.271.0–10.40.0450.790.23–2.70.711
 ≤0.10 vs >0.105.592.0–15.30.0011.750.59–5.10.311
 ≤0.15 vs >0.157.092.6–19.4<0.0012.280.78–6.70.133
 ≤0.25 vs >0.256.472.8–14.9<0.0012.801.1–7.00.027
 ≤0.5 vs >0.57.153.7–14.0<0.0012.341.1–5.20.037
 ≤1.0 vs >1.05.623.4–9.4<0.0011.800.95–3.40.071
 ≤2.0 vs >2.04.292.7–6.8<0.0011.170.63–2.20.619
 ≤3.0 vs >3.03.172.0–5.1<0.0010.810.44–1.50.812
CRPContinuous1.111.1–1.1<0.0011.040.98–1.10.174
Table 3. Univariable model of possible independent prognostic variables in patients with clear cell RCC with a categorized CRP level of 0.25 mg/dL
CovariableCategoriesUnivariable
HR95% CIPPredictive accuracy (%)
CRP≤0.25 vs >0.256.472.8–14.9<0.00164.8
T stageT1/2 vs T3/46.864.1–11.5<0.00173.2
N stageN0/X vs N1/27.794.3–14.2<0.00159.1
M stageM0 vs M113.408.3–21.8<0.00173.4
Nuclear gradeG1/2 vs G3/49.025.6–14.5<0.00171.5
Karnofsky index≥80% vs <80%1.961.2–3.20.00658.8
Tumour size≤7 vs >7 cm4.472.8–7.2<0.00165.4
Table 4. Five-year overall and cancer-specific survival
 5-year survival (%) (95% CI)
Overall survivalCancer-specific survival
  1. a

    Kaplan–Meier survival analyses, P < 0.001.

All patients65.1 (62.3–67.9)78.1 (74.6–80.6)
CRP  
≤0.25 mg/dL82.1 (78.1–86.1)a92.3 (89.1–95.5)a
>0.25 mg/dL56.1 (59.6–52.5)a69.9 (66.5–73.3)

As CRP 0.25 mg/dL proved to be the best independent prognostic variable to determine a CRP threshold, predictive accuracy was calculated with and without the inclusion of CRP 0.25 mg/dL. In the full CP base model, including the traditional predictor variables of TNM stage, grading, Karnofsky status performance scale and tumour size, predictive accuracy was 87.8%; with the addition of CRP 0.25 mg/dL predictive accuracy was 87.9%. In a second step we reduced the CP base model by variables that were not significant in the full multivariable Cox model (tumour size excluded). The predictive accuracy of the reduced CP base model was 88.1%; the addition of CRP 0.25 mg/dL to the reduced CP base model resulted in no additional increase of the predictive accuracy (87.7%, Table 5).

Table 5. Multivariable model of possible independent prognostic variables in patients with clear cell RCC with a categorized CRP level of 0.25 mg/dL
CovariableCategoriesFull multivariable base modelFull multivariable model, CRP includedReduced multivariable base modelReduced multivariable model, CRP included
HR95% CIPHR95% CIPHR95% CIPHR95% CIP
CRP≤0.25 vs >0.25   2.801.1–7.00.027   2.651.1–6.60.036
T stageT1/2 vs T3/42.921.5–5.80.0032.441.2–4.90.0132.561.3–4.90.0042.151.1–4.10.023
N stageN0/X vs N1/22.751.4–5.40.0032.581.3–5.00.0052.381.3–4.50.0062.211.2–4.10.013
M stageM0 vs M16.653.8–11.5<0.0016.913.9–12.2<0.0016.413.8–11.0<0.0016.523.8–11.3<0.001
Nuclear gradeG1/2 vs G3/44.572.5–8.5<0.0014.052.2–7.5<0.0014.102.3–7.3<0.0013.562.0–6.4<0.001
Karnofsky index≥80% vs <80%2.141.3–3.60.0042.141.3–3.60.0042.171.3–3.60.0032.171.3–3.60.003
Tumour size≤7 vs >7 cm0.650.35–1.20.1770.600.33–1.10.105      
Predictive accuracy (%) 87.8  87.9  88.1  87.7  

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References

WBC count is an easy and quick marker to detect a host's response to inflammation. Besides the wide use of WBC count, serum CRP represents a sensitive indicator of inflammatory activity. CRP was the first member of a group of proteins identified as ‘acute phase proteins’ which can serve as indicatory serum proteins in early inflammation and infection, if they are elevated [24]. Here, CRP is a more sensitive marker of early inflammation than WBC count. Today, CRP and WBC count can be measured by standardized tests, and they are routinely used in the clinic to screen for acute or chronic inflammation.

A systemic inflammatory response reflected by the elevation of WBC count and acute phase proteins is also detected in various cancers [9]. Systemic inflammation has been linked with the progression of urological cancers, and it was associated with a limited prognosis in RCC and urothelial and prostate cancer [25]. On the one hand inflammation and activation of the immune system enable anti-tumour activity, on the other hand they contribute to carcinogenesis, tumour growth and progression in human cancers [26]. Local inflammation is reflected by intratumoural infiltration of leucocytes, macrophages and neutrophils, which were identified as an independent factor for reduced survival in clear cell RCC [27]. Besides the prognostic factors of high lactate dehydrogenase, lymph node metastases, low haemoglobin or low Karnofsky performance status, an elevated WBC count of >6.0/μL, the intratumoural invasions of neutrophils and a low intratumoural subset of CD57 positive natural killer cells are independent poor prognostic immunological factors [19]. This analysis confirmed the prognostic differences of different WBC thresholds, ranging from 9.5 to 11/μL.

Moreover, the prognostic relevance of preoperatively analysed inflammatory biomarkers has been reported for patients with RCC with curative intended surgery as well as cytoreductive surgery in the metastatic setting [19, 28-30]. But, for a better prediction of recurrence and mortality the use of postoperative biomarker measurement or the determination of kinetics has been postulated by some investigators [17, 31, 32]. However, in RCC postoperative biomarker methods are biased due to significant differences in short- and long-term related tissue trauma. The different surgical techniques of an open vs a laparoscopic approach or nephron sparing surgery vs radical nephrectomy and the reduction of renal function results in a variation in the levels and kinetics of any biomarker. For example, significantly increased CRP plasma levels are measured after nephrectomy of large tumours due to a reduced renal function [33, 34]. Therefore, for better reliability we performed standardized WBC count and CRP measurements in the preoperative setting. A large number of studies have confirmed the independent prognostic significance of CRP measured before kidney surgery. All these studies showed CRP to be of prognostic value, but taken together the CRP concentrations of the studies showed a range from 0.4 to 2.3 mg/dL and none of the studies determined a potential threshold value to be the best [28-30].

In a study of 286 patients with RCC with a median follow-up of 5 years a CRP ≥ 1.5 mg/dL was a significant predictor of overall and cancer-specific survival [29]. Recently, in a study of 451 T1a tumours a CRP value of ≥ 0.4 mg/dL was an independent prognostic factor for metastasis and overall survival [30]. In a meta-analysis including 18 different CRP studies (n = 2927) values ranged from 0.3 to 1.1 mg/dL. In these studies patients with RCC with an increased preoperative CRP had a higher risk of cancer-specific death (HR 3.46, 95% CI 2.80–4.27) [14].

Other reports have proposed a range of threshold levels for WBC count, which varied between 7.6 and 11.0/nL [29, 32, 35]. Therefore a model incorporating a WBC value >7.6/nL was introduced by the International Kidney Cancer Working Group together with different parameters of treatment modality, performance status, number of metastases, time to treatment, lactate dehydrogenase, alkaline phosphatase and serum calcium to improve the predictive accuracy [32]. The number of studies using both WBC count and CRP as prognostic markers is limited, but for both variables a prediction of cancer-specific survival and overall survival in patients undergoing potentially curative or cytoreductive nephrectomy was shown [29].

Although the prognostic value of CRP and WBC count as biomarkers seems clearly evident, there are major drawbacks regarding the routine use of inflammatory biomarkers compared with other cancer-specific prognostic factors. First, WBC count and CRP are non-specific serum biomarkers indicating local or systemic inflammation. Systemic inflammation plays an important role in the pathogenesis of atherosclerosis and was identified as an independent risk factor for cardiovascular mortality in patients with chronic kidney disease [36, 37]. Second, no clear threshold levels are defined to identify patients with a higher likelihood of disease progression and reduced survival. Third, although CRP and WBC count are routinely analysed before surgery in almost all patients with RCC, only a few reports have made use of both biomarkers for their prognostic stratifications.

Therefore the aim of the study was to re-evaluate the prognostic significance of preoperative WBC count and CRP in one large cohort of patients with clear cell RCC that independently predicts patient prognosis beyond other accepted factors. In a second step optimal threshold values for WBC count and CRP were determined. The analysis of WBC count could detect 9.5/nL and 11.0/nL as potential discriminatory values, but a clear breakpoint was not visible. In contrast, univariable analysis of different CRP threshold values revealed an increase of the HR up to 6.47 (CRP 0.25 mg/dL) and 7.15 (CRP 0.50 mg/dL) with a decline thereafter, identifying a potential CRP breakpoint in this range. The adjustment for TNM stage, Fuhrman nuclear grade, Karnofsky index and tumour size in the multivariate Cox proportional hazard model showed a CRP value of 0.25 mg/dL to be best for the differentiation between the groups (≤0.25 vs >0.25, HR 2.80, 95% CI 1.1–7.0, P = 0.027).

Interestingly, only patients with the presence of metastatic disease (HR 6.91) and a high Fuhrman grade (HR 4.05) were at higher risk for cancer-specific death than the group of patients with a CRP of ≤0.25 vs >0.25 mg/dL (HR 2.80). In accordance with these results patients with a higher CRP value (>0.25 mg/dL) were associated with a higher tumour grading, higher tumour stage and with the presence of distant metastasis and regional lymph node metastasis.

Although the connection between inflammation and cancer seems evident, the exact mechanism between increased CRP and WBC count and tumour progression is not clearly resolved. The different types of leucocytes (Th1, Th2, regulatory T cells) which are not further differentiated by a general WBC count contribute to the missing robustness of this parameter in the multivariate model. Systemic immune response of the host to systemic tumour growth, local tissue inflammation induced by tumour infiltration or production of pro-inflammatory proteins by tumour cells are only some of several possible explanations for high CRP values in patients with higher stages of clear cell RCC [19]. This link between cancer and systemic inflammatory response parameters enables clinicians to improve the accuracy of established prognostic algorithms and was introduced in recent prediction models for urological cancers [25].

Despite the independence of CRP 0.25 mg/dL in the multivariable Cox model, CRP was not able to raise predictive accuracy in this cohort of patients. The CP base model, which included the traditional predictor variables of TNM stage, grading and Karnofsky performance status, was of a high predictive accuracy (88.1%) which could not be further improved by the addition of CRP (87.1%). In this study the predictive accuracy (c-index) of the CP base model is higher than it was in other studies which explored CRP as a biomarker in RCC. A high predictive accuracy of a base model is hard to improve by additional variables. For example, Iimura et al. analysed 249 patients with RCC, and the c-index of their base model (TNM stage) was 0.792; if CRP was added to the base model the c-index was 0.820 [18]. On the other hand, in a large study by Karakiewicz et al. the traditional predictor variables (TNM stage, grading, symptoms at presentation) demonstrated a high predictive accuracy of 86.3% in patients with RCC [38]. In our opinion the additional predictive value of a biomarker is somehow limited if the underlying base model already shows high predictive accuracy as was the case in the present cohort of patients.

The open question is whether inflammatory biomarkers will become preoperative predictive factors, or if they may serve as markers in next-generation drug and specific immune therapy regimes as was demonstrated in a prospective phase I/II study for the biomarkers apolipoprotein A-I and chemokine CCL17 in patients with metastatic RCC treated with a peptide vaccine [39].

In conclusion, this study did not exclude biased values of CRP and WBC count due to local or smouldering chronic infection, but significantly confirmed CRP to be an independent prognostic parameter in clear cell RCC. The breakpoint analysis shows a CRP value of 0.25 mg/dL to be the best to differentiate the prognosis between the groups. Nevertheless CRP 0.25 mg/dL could not improve predictive accuracy in this cohort of patients. Currently the clinical use of CRP is not necessary if traditional parameters like TNM stage, grading and Karnofsky index are carefully assessed.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Conflict of Interest
  8. References
  • 1
    Siegel R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: the impact of eliminating socioeconomic and racial disparities on premature cancer deaths. CA Cancer J Clin 2011; 61: 212236
  • 2
    Kruck S, Bedke J, Kuczyk MA, Merseburger AS. Second line systemic therapy for the treatment of metastatic renal cell cancer. Expert Rev Anticancer Ther 2012; 12: 777785
  • 3
    Karakiewicz PI, Suardi N, Capitanio U et al. A preoperative prognostic model for patients treated with nephrectomy for renal cell carcinoma. Eur Urol 2009; 55: 287295
  • 4
    Bedke J, Buse S, Pritsch M et al. Perinephric and renal sinus fat infiltration in pT3a renal cell carcinoma: possible prognostic differences. BJU Int 2009; 103: 13491354
  • 5
    Ljungberg B, Cowan NC, Hanbury DC et al. EAU guidelines on renal cell carcinoma: the 2010 update. Eur Urol 2010; 58: 398406
  • 6
    Nguyen MM, Gill IS. Effect of renal cancer size on the prevalence of metastasis at diagnosis and mortality. J Urol 2009; 181: 10201027
  • 7
    Maxine S, Shahrokh FS, Christopher C et al. Prognostic factors and predictive models in renal cell carcinoma: a contemporary review. Eur Urol 2011; 60: 644661
  • 8
    Kruck S, Merseburger AS, Hennenlotter J et al. High cytoplasmic expression of p27(Kip1) is associated with a worse cancer-specific survival in clear cell renal cell carcinoma. BJU Int 2012; 109: 15651570
  • 9
    Balkwill F, Mantovani A. Inflammation and cancer: back to Virchow? Lancet 2001; 357: 539545
  • 10
    Kataoka H, Tanaka H, Nagaike K, Uchiyama S, Itoh H. Role of cancer cell–stroma interaction in invasive growth of cancer cells. Hum Cell 2003; 16: 114
  • 11
    Bedke J, Stenzl A. Immunologic mechanisms in RCC and allogeneic renal transplant rejection. Nat Rev Urol 2010; 7: 339347
  • 12
    Bedke J, Laske K, Feyerabend S et al. 76 Clinical and immunological monitoring of a multi-peptide vaccination trial in patients with stage III renal cell carcinoma. Eur Urol 2012; 11: e76ea
  • 13
    Whiteside T. The role of immune cells in the tumor microenvironment the link between inflammation and cancer. In Dalgleish AG , Haefner B eds, The Link Between Inflammation and Cancer: Wounds That Do Not Heal, Vol. 130. New York: Springer, 2006: 103124
  • 14
    Wu Y, Fu X, Zhu X et al. Prognostic role of systemic inflammatory response in renal cell carcinoma: a systematic review and meta-analysis. J Cancer Res Clin Oncol 2011; 137: 887896
  • 15
    Gakis G, Todenhofer T, Renninger M et al. Development of a new outcome prediction model in carcinoma invading the bladder based on preoperative serum C-reactive protein and standard pathological risk factors: the TNR-C score. BJU Int 2011; 108: 18001805
  • 16
    Wang CS, Sun CF. C-reactive protein and malignancy: clinico-pathological association and therapeutic implication. Chang Gung Med J 2009; 32: 471482
  • 17
    Johnson TV, Abbasi A, Owen-Smith A et al. Postoperative better than preoperative C-reactive protein at predicting outcome after potentially curative nephrectomy for renal cell carcinoma. Urology 2010; 76: 766 e1–5
  • 18
    Iimura Y, Saito K, Fujii Y et al. Development and external validation of a new outcome prediction model for patients with clear cell renal cell carcinoma treated with nephrectomy based on preoperative serum C-reactive protein and TNM classification: the TNM-C score. J Urol 2009; 181: 10041012
  • 19
    Ramsey S, Lamb GW, Aitchison M, Graham J, McMillan DC. Evaluation of an inflammation-based prognostic score in patients with metastatic renal cancer. Cancer 2007; 109: 205212
  • 20
    Negrier S, Gomez F, Douillard JY et al. Prognostic factors of response or failure of treatment in patients with metastatic renal carcinomas treated by cytokines: a report from the Groupe Francais d'Immunotherapie. World J Urol 2005; 23: 161165
  • 21
    Fuhrman SA, Lasky LC, Limas C. Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982; 6: 655663
  • 22
    Harrell FE Jr, Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the yield of medical tests. JAMA 1982; 247: 25432546
  • 23
    Taylor JM, Ankerst DP, Andridge RR. Validation of biomarker-based risk prediction models. Clin Cancer Res 2008; 14: 59775983
  • 24
    Tillett WS, Francis T. Serological reactions in pneumonia with a non-protein somatic fraction of pneumococcus. J Exp Med 1930; 52: 561571
  • 25
    Saito K, Kihara K. C-reactive protein as a biomarker for urological cancers. Nat Rev Urol 2011; 8: 659666
  • 26
    Coussens LM, Werb Z. Inflammation and cancer. Nature 2002; 420: 860867
  • 27
    Jensen HK, Donskov F, Marcussen N, Nordsmark M, Lundbeck F, von der Maase H. Presence of intratumoral neutrophils is an independent prognostic factor in localized renal cell carcinoma. J Clin Oncol 2009; 27: 47094717
  • 28
    Karakiewicz PI, Hutterer GC, Trinh QD et al. C-reactive protein is an informative predictor of renal cell carcinoma-specific mortality: a European study of 313 patients. Cancer 2007; 110: 12411247
  • 29
    Jagdev SP, Gregory W, Vasudev NS et al. Improving the accuracy of pre-operative survival prediction in renal cell carcinoma with C-reactive protein. Br J Cancer 2010; 103: 16491656
  • 30
    Takayama T, Sugiyama T, Kai F et al. Characteristics of aggressive variants in T1a renal cell carcinoma. J Cancer Res Clin Oncol 2011; 137: 16531659
  • 31
    Saito K, Tatokoro M, Fujii Y et al. Impact of C-reactive protein kinetics on survival of patients with metastatic renal cell carcinoma. Eur Urol 2009; 55: 11451153
  • 32
    Manola J, Royston P, Elson P et al. Prognostic model for survival in patients with metastatic renal cell carcinoma: results from the International Kidney Cancer Working Group. Clin Cancer Res 2011; 17: 54435450
  • 33
    Panichi V, Migliori M, De Pietro S et al. C reactive protein in patients with chronic renal diseases. Ren Fail 2001; 23: 551562
  • 34
    Shlipak MG, Fried LF, Crump C et al. Elevations of inflammatory and procoagulant biomarkers in elderly persons with renal insufficiency. Circulation 2003; 107: 8792
  • 35
    Ramsey S, Lamb GW, Aitchison M, McMillan DC. Prospective study of the relationship between the systemic inflammatory response, prognostic scoring systems and relapse-free and cancer-specific survival in patients undergoing potentially curative resection for renal cancer. BJU Int 2008; 101: 959963
  • 36
    Jialal I, Devaraj S, Venugopal SK. C-reactive protein: risk marker or mediator in atherothrombosis? Hypertension 2004; 44: 611
  • 37
    Menon V, Greene T, Wang X et al. C-reactive protein and albumin as predictors of all-cause and cardiovascular mortality in chronic kidney disease. Kidney Int 2005; 68: 766772
  • 38
    Karakiewicz PI, Briganti A, Chun FK et al. Multi-institutional validation of a new renal cancer-specific survival nomogram. J Clin Oncol 2007; 25: 13161322
  • 39
    Walter S, Weinschenk T, Stenzl A et al. Prolonged survival of patients with advanced renal cancer responding to multi-peptide vaccine IMA901 after single-dose cyclophosphamide. Nat Med 2012; 18: 12541261
Abbreviations
WBC

white blood cell

CRP

C-reactive protein

HR

hazard ratio

CP

clinicopathological