Role of C-reactive protein in urological cancers: A useful biomarker for predicting outcomes

Authors


Kazutaka Saito M.D., Ph.D., Department of Urology, Graduate School of Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan. Email: kz-saito.uro@tmd.ac.jp

Abstract

Based on increasing evidence of the association between cancer-related inflammation and the progression of cancer, the external symptoms of systemic inflammatory response has been shown to be an indicator for the prognosis of many malignancies, including urological cancers. C-reactive protein, a representative acute-phase reactant, is a significant and sensitive inflammatory marker that can be objectively measured using reliable assays in clinical practice worldwide. C-reactive protein has been shown to be significant in the prediction of outcomes of urological cancers. The elevation of C-reactive protein levels, which indicate the presence of cancer-associated systemic inflammatory response, is linked to poorer survival in patients with urological cancers, including renal cell carcinoma, upper urinary tract and bladder cancers, and prostate cancer. With this strong prognostic ability, C-reactive protein can be incorporated into prognostic models and will make them simpler and improve their predictive accuracy. Furthermore, the longitudinal change of C-reactive protein level, C-reactive protein kinetics, provides additional information on patient survival outcomes. As such, C-reactive protein can be used to monitor treatment efficacy and disease course using serial measurements. In testicular cancer, C-reactive protein is associated with a risk of late complications, such as cardiovascular disease, and with the development of second non-germ-cell cancer. Taken together, these findings show that C-reactive protein can act as an important biomarker for urological cancers. This review discusses the importance of C-reactive protein as a prognostic biomarker in urological cancers on the basis of the currently available evidence.

Abbreviations & Acronyms
C/EBP =

CCAAT enhancer binding protein

CRP =

C-reactive protein

CRPC =

castration-resistant prostate cancer

CSS =

cancer-specific survival

GPS =

Glasgow Prognostic Score

HR =

hazard ratio

IFN =

interferon

IL =

interleukin

MIBC =

muscle invasive bladder cancer

MRCCPS =

Metastatic Renal Carcinoma Comprehensive Prognostic System

MSKCC =

Memorial Sloan-Kettering Cancer Center

NIH =

National Institutes of Health

OS =

overall survival

PFS =

progression-free survival

PSA =

prostate-specific antigen

RCC =

renal cell carcinoma

RFS =

recurrence-free survival

TNM =

tumor–nodes–metastasis

UC =

urothelial carcinoma

UUT =

upper urinary tract

Introduction

The seeking of biomarkers is essential to improving the management of urological cancers. According to the 2001 USA NIH definition, a novel biomarker can be proposed if it has a characteristic that can be measured and evaluated as an indicator of normal biological processes, pathogenic processes or pharmacological responses to a therapeutic intervention.1 For cancer, a biomarker candidate will be judged by its prognostic impact in predicting the likelihood of metastasis or survival, in judging diagnostic and staging tools, as well as its ability to predict and monitor the clinical response to an intervention. The practical application of a specific biomarker would allow us to make precise risk assessment of disease progression and to implement improved therapeutic approaches, such as future clinical trials and simple surveillance protocols related to individual risk.

Along with the TNM classification and histological grade of tumors, which have been used as the main prognostic indicators in urological cancers, other specific prognostic factors have been explored in order to identify biomarkers. As cancer-related features that are linked to disease outcomes reflect the underlying mechanism of disease progression, the identification of a biomarker might lead to the development of a novel therapeutic approach directly against the related tumor features.

Among the various cancer-related features, inflammation-associated factors have been considered as potential biomarkers, based on the evidence that underlying inflammatory processes play important roles in cancer progression. The presence of systemic inflammation indicated by levels of acute-phase reactant proteins has been shown to predict poor prognosis in various cancers. Among the inflammatory markers, CRP, a representative acute-phase reactant, has a significant impact on the prognosis of many cancers, including urological cancers. Furthermore, the CRP kinetics of dynamic changes in CRP levels can be used to monitor the disease course, such as the effect of treatment intervention or further progression. CRP could thus serve as a useful biomarker for cancer.

In the present review, the role of CRP and systemic inflammation in the progression of cancer will be described, and we will discuss the potential of CRP as a useful biomarker based on the presented evidence in urological cancers, including RCC, UC, prostate cancer and testicular cancer.

Systemic inflammation and CRP in cancer

Systemic inflammation is a complex biological response in vascular tissues that is caused by harmful stimuli, such as pathogens, irritants or damaged cells.2 Inflammation is not synonymous with infection, because inflammation is a protective response by the host to remove harmful stimuli and to prompt tissue healing. In fact, a number of disorders, including immune disorders and cardiovascular disease, can cause inflammation. Furthermore, the inflammatory process is involved in the pathogenesis of various disorders.

The inflammatory process also involves the underlying mechanisms that relate to cancer progression. Cancer-associated inflammation plays an important role in various stages of tumor development, including disease progression, malignant conversion, invasion and metastasis.3 A tumor can trigger an inflammatory response (cancer-associated inflammation), which leads to the production of major inflammatory cytokines, including IL-1β and IL-6.4 Cancer progression can be induced in an autocrine or paracrine manner through the aberrant expression of inflammatory cytokines by cancer cells.5–8 The cancer-associated reaction of the systemic inflammatory response in the host is found throughout the entire organ system, including many aspects of neuroendocrine metabolism, hematopoietic function, and protein and energy metabolism.9 Recent investigations have revealed that external symptoms of systemic inflammation are significant predictors of poor outcome in many cancers.10–13 To obtain a better understanding of the underlying mechanism of inflammation as it relates to cancer progression, some therapeutic approaches that target the cancer-associated inflammation have been shown to be effective, such as cyclooxygenase-2 inhibitor in combination with IFN-α for RCC,14,15 and aspirin for colorectal cancer.16

Quantification of the systemic inflammatory response is essential to adequately evaluate the activity or outcomes of inflammation-related diseases. The severity of the systemic inflammation is usually assessed by the level of acute-phase reactant. Among them, CRP is a representative acute-phase protein that can serve as a definitive marker for systemic inflammation.

CRP was first discovered in 1930 in sera obtained from patients in the acute phase of pneumococcal pneumonia. It was named for its reaction to the cell wall of pneumococcal C polysaccharide.17 CRP is a member of the pentraxin family of proteins (with five protomers around a central pore) and belongs to the short pentraxin subtype of the superfamily.18 CRP is an element of the innate immune system.19 A major role of CRP is to bind phosphocholine, thereby permitting recognition of both foreign pathogens and phospholipid constituents of damaged cells.20 CRP also plays a role in host defense, and in the clearance of necrotic and apoptotic cells.20

In humans, serum CRP levels can be markedly elevated in the early phase as a result of inflammatory stimulus. CRP is primarily synthesized in hepatocytes at the transcriptional level under stimulation mainly by IL-6 and IL-1β, which are produced in various cells, including inflammatory cells, such as T cells and macrophages, endothelial cells, fibroblasts and some cancer cells,21 although this mechanism is not fully understood.22 The CRP gene transcription is mainly activated by C/EBP family members through the stimuli of these cytokines.23 The synthesis of CRP at sites other than the liver has also been reported in neurons and atherosclerotic plaque tissue.24,25 The mechanisms regulating the extrahepatic synthesis of CRP remain to be determined, but extrahepatic sources are not considered to contribute to plasma CRP levels.22 Two studies have, however, shown that some RCC cells can produce CRP and contribute to serum CRP levels.26,27

Serum CRP measurement has been established with reliable assays in clinical practice. Population-based studies have shown a skewed distribution of CRP levels. In the reference population, approximately 70–90% of samples have a CRP concentration of less than 3 mg/L, although a slight elevation in CRP levels of up to 10 mg/L was found in some individuals.28 Infection, most often bacterial, causes marked CRP elevation of more than 100 mg/L, which is found in approximately 80% of patients.29 In contrast, minor elevation of CRP levels between 3 and 10 mg/L might reflect a physiological condition of minor inflammation or a response to metabolic stress, such as obesity, cigarette smoking, diabetes mellitus or hypertension.30 Recently, high-sensitivity CRP assay that can detect low levels of CRP (below 1 mg/L) has been used to estimate the risk of cardiovascular disease.31

CRP level has been shown to be prognostic in various inflammation-related disorders.32 For instance, CRP is a predictor for the onset of rheumatoid arthritis.33 In cardiovascular disease, CRP level is associated with the long-term risk of a myocardial infarction, ischemic stroke or peripheral vascular disease.34–36

In cancer patients, serum CRP level shows elevation. In a study of 496 cases and 996 controls, the serum CRP levels in cancer cases were higher than those in controls.37 On the basis of the association between the presence of systemic inflammation and poorer outcome, CRP has been shown to be an important predictor of survival in patients with various cancers, including colon, lung, breast and ovarian cancer, among others.9 As in other inflammation-related disease, but not bacterial infection, the cut-off point for the elevated CRP level in malignancy was set at 3–10 mg/L in the range of minor CRP elevations.9 As CRP is a non-specific inflammatory marker that value increases by other inflammation-related disease including infection, we should evaluate the CRP values in cancer patients, paying attention to the possibility of other inflammation-related disease that could affect those. CRP levels along with other systemic inflammation markers, such as white cell count and its components, provide input for prognostic algorithms, such as systemic-inflammation-based prognostic scores.38 The importance of CRP as an accurate and readily available biomarker in urological cancers is discussed in the present review.

Prognostic value of CRP in renal cell carcinoma

In urological cancer, the importance of CRP as a potential biomarker has been investigated most intensively in RCC.39 Numerous studies have shown that CRP is a significant prognostic factor for RCC patients treated with surgery and/or systemic therapy (Table 1).

Table 1. Prognostic significance of CRP in RCC
StudyPatientsStageTreatmentCut-off of CRP (mg/L)Prognostic value of CRP
Masuda et al.40111TanyNanyM0NephrectomyNACRP and T stage are important prognostic factors for OS.
Lamb et al.41100TanyN0M0Nephrectomy10CRP is an independent prognostic factor for CSS.
Komai et al.42101T1-3N0M0Nephrectomy5CRP is an independent prognostic factor for PFS and CSS.
Ramsay et al.4383cT1-3N0M0Nephrectomy10A prospective study reveals that elevated CRP predicts recurrence.
Johnson et al.44130T1-3N0M0Nephrectomy• (Absolute CRP)CRP predicts 1-year RFS and OS.
Ito et al.45178AnyNephrectomy10CRP is a prognostic factor for RFS and CSS.
Karakiewicz et al.46314AnyNephrectomy4, 23CRP improves the predictive accuracy for CSS.
Iimura et al.47539AnyNephrectomy5CRP is combined into a new prognostic model with TNM stage (TNM-C score) to predict CSS.
Johnson et al.48110T1-3N0M0Nephrectomy• (Absolute CRP)Postoperative CRP is superior to predict prognosis compared with preoperative CRP.
Ito et al.49263TanyN0M0Nephrectomy10 (Preoperative)Postoperative C-reactive protein non-normalization is an independent predictors of OS
3 (Postoperative)
Tatokoro et al.5040MetastaticCytoreductive nephrectomy5 (Pre- and postoperative)Preoperative CRP kinetics predicts OS. Non-normalization of CRP after nephrectomy indicated poor prognosis.
Bromwich et al.5158Cytokine10CRP predicts poor survival.
Casamassima et al.52110MetastaticCytokine8CRP predicts survival of patients treated with IL-2.
Vogl et al.5399MetastaticCytokine8CRP predicts OS.
Atzpodien et al.54425MetastaticCytokine11CRP can be combined into a prognostic scoring system to predict OS.
Ramsay et al.55119MetastaticCytokine10A prognostic score with CRP and albumin (GPS system) clearly stratify OS.
Shinohara et al.1443MetastaticCytokine4CRP kinetics is associated with PFS.
Saito et al.56108MetastaticMultimodality5CRP kinetic predicts OS. Normalized CRP period can directly predict overall survival period.

To the best our knowledge, the significant association between the presence of systemic inflammatory response, including elevated CRP levels, and poor prognosis in RCC patients treated by nephrectomy was first shown in 1973 in a Japanese-language article by Satomi.57 Satomi mentioned that the postoperative normalization of CRP indicated better prognosis compared with the non-normalization of postoperative CRP. These findings were not confirmed or re-evaluated until the last decade.

Masuda et al. reported the prognostic value of CRP in 111 RCC patients who underwent nephrectomy in 1998.40 Ito et al. reported that elevated CRP levels (10 mg/L or greater) were associated with poor CSS in 178 RCC patients.45 Our first study with 101 pT1-3N0M0 RCC patients showed that elevated CRP levels of 5 mg/L or greater, as well as pathological T3 stage, predicted poorer RFS and CSS. The 5-year RFS and CSS rates in patients with elevated CRP levels were 55.8 and 75.4%, respectively, compared with 93.1 and 94.8%, respectively, in patients without CRP elevation (P < 0.0001).42 Lambs et al. observed that elevation of CRP levels (>10 mg/L) was linked to poorer CSS in 100 RCC patients without metastasis.41 Thereafter, they showed the significance of CRP elevation (>10 mg/L) to predict poorer RFS and CSS in a prospective study with 83 RCC patients undergoing curative nephrectomy.43 Absolute preoperative CRP as a continuous variable is also a good predictor of metastasis and mortality 1 year after surgery in clinically localized RCC patients.44

The prognostic significance of CRP has been shown with its power for improving the combined predictive accuracy of established markers and prediction models through multivariable statistics46,47 in a manner consistent with a criterion proposed by Kattan.58 In a study by Karakiewicz et al. that included 314 patients with any stage of RCC, CRP improved the predictive accuracy by 3.7% for RCC-specific mortality.

In addition to nephrectomized cases, CRP is also a prognostic and predictive marker for metastatic RCC patients receiving systemic cytokine therapy. CRP can predict response to cytokine therapies (IFN-α and/or IL-2) with an inverse correlation between CRP level and response. Patients with high CRP concentrations (>50 mg/L) had an increased risk of cancer progression during IL-2 therapy than those without CRP elevations.59 Miyake et al. showed that pretreatment CRP levels had a significant impact on the response to receiving IL-2 and INF-α combination therapy.60 Although the response rate itself is not considered a good surrogate for OS in metastatic RCC,61 CRP is not only a predictive marker for response, but also a significant prognostic indicator for survival in patients with metastatic RCC.51–53,55,62 In a large Japanese multicenter cohort of 1463 metastatic RCC patients,62 CRP had significant prognostic impact on OS, as well as time from initial visit to metastasis, Eastern Cooperative Oncology Group Performance Status, serum lactate dehydrogenase and Ca level, all of which were included in the MSKCC criteria.63 The 1-year OS rate of 598 patients with elevated CRP levels (>3 mg/L) was 55%, compared with 84.7% of 266 patients without those. One study showed that pretreatment CRP levels had prognostic value for both patients who responded to IL-2 therapy and those with stable or progressive disease.52 Therefore, CRP might be more suitable as a surrogate for survival than response rate.

Incorporation of CRP into prognostic algorithms in RCC patients

Based on the strong impact of CRP in predicting the survival of RCC patients, the incorporation of CRP into prognostic algorithms has been considered. Two prognostic algorithms have been developed for application to metastatic RCC: the MRCCPS54 and the GPS.55 In the GPS, both elevated CRP levels (>10 mg/L) and hypoalbuminemia (<35 g/L) were given a score of 2, whereas either elevated CRP or hypoalbuminemia were given a score of 1. Albumin is a negative acute-phase reactant whose levels decreased under systemic inflammation. The superiority of the GPS's predictive accuracy was observed in comparison with that of the MSKCC criteria and the MRCCPS.

We developed and validated a simple combined scoring system, the TNM-C score, to predict CSS of all-stage clear-cell RCC patients who had undergone nephrectomy.47 The score consists solely of the presence or absence of elevated CRP levels and TNM classification; tumor stage pT3 was given a score of 2, presence of regional lymph node metastasis (N1 or N2) was given a score of 2, presence of distant metastasis (M1) was given a score of 4, and CRP concentration was given a score of 1 (if ≥5 mg/L) or 0 (if <5 mg/L). The addition of CRP data into the prognostic algorithm improved the model's predictive accuracy by 6%. The predictive accuracy of the score in the model development cohort was 0.846, as obtained by the c index that was externally validated with a predictive accuracy of 0.876. Despite its simplicity, the predictive accuracy remained good. Usually, detailed pathological findings, such as nuclear grade and tumor necrosis, are required in predictive algorithms,64 although the standardization of nuclear grade criteria is required to improve interobserver reproducibility65 and the evaluation of tumor necrosis is not common.66 With the strong correlation between CRP and tumor aggressiveness features determined by pathological examination, CRP could represent them in prognostic algorithms.

CRP kinetics in renal cell carcinoma

To show the usefulness of CRP in the prediction and monitoring of clinical response to a therapeutic intervention, the impact of post- and/or intratreatment CRP level on disease outcome can be assessed. In a study by Johnson et al., postoperative CRP predicted the prognosis of localized RCC patients undergoing curative nephrectomy more significantly than preoperative CRP.48 We showed that perioperative dynamic changes in CRP concentration, CRP kinetics, predicted the clinical course of metastatic RCC patients treated with cytoreductive nephrectomy.50 In this study, patients were divided into three groups based on their preoperative CRP levels and the status of CRP normalization 1 month after the operation. Patients whose preoperative CRP was not elevated (CRP <5 mg/L) were assigned to the non-elevated CRP group. The remaining patients with preoperative CRP ≥5 mg/L were further divided into two groups according to postoperative CRP normalization (<5 mg/L) or not, and assigned to the normalized CRP group or non-normalized CRP group, respectively. Adding to the baseline status, the failure of CRP to normalize after surgery was associated with extremely poor prognosis. The median OS period of the non-elevated, normalized and non-normalized CRP groups was 30, 21 and 5 months, respectively. Ito et al. also showed the prognostic impact of the baseline CRP status (10 mg/L) and postoperative CRP normalization (<3 mg/L) in 263 localized RCC patients.49 Taken together, these findings show that perioperative CRP kinetics might add useful information to the prediction of outcome for RCC patients treated surgically.

Shinohara et al. also showed that CRP kinetics with patient stratification based on baseline and nadir CRP level during treatment (cut-off point: 4 mg/L) was associated with PFS in 36 patients with metastatic RCC who were treated with IFN-α combination therapy.14 We provided evidence supporting the use of CRP kinetics as a dynamic marker of survival in patients with metastatic RCC.56 Patients treated with multimodal therapy, including nephrectomy, metastasectomy and cytokine therapy, were grouped according to CRP kinetics with baseline and nadir CRP levels (cut-off point: 5 mg/L) as non-elevated, normalized and non-normalized CRP groups. CRP kinetics predicted OS, improving the predictive accuracy by 4% to baseline CRP alone. Survival curves were clearly stratified by CRP kinetics with 2-year OS rates of 69, 55 and 4%, and median survival of 45 months, 28 months and 5 months for the non-elevated, normalized and non-normalized groups, respectively. Furthermore, the study showed that the serial measurement of CRP might be useful for the estimation of OS length in individual patients. The normalized CRP period, namely, the duration of normal CRP level without persistent elevation, was associated with OS period, and a patient's OS time could be estimated at approximately 1 year plus the length of the normalized CRP period. On the basis of this evidence, we can evaluate the treatment efficacy and expected length of survival with serial CRP levels. Non-elevation or decrease to the normal range of CRP might indicate relatively long survival periods. In contrast, a continuous increase in CRP levels was linked to disease progression. We previously reported a case in which a rare RCC metastasis to the stomach was found within the operable state as a result of the continuous increase in CRP levels.67

The aforementioned results show that CRP kinetics can be used to estimate the risk of disease progression to death. Taken together, these findings indicate that CRP is an important biomarker in RCC that satisfies the NIH biomarker criteria.1 CRP is a significant indicator of normal or pathological processes that can be used to predict the outcome of RCC. In addition, serial measurements of CRP concentration can be used to monitor the response to therapy or disease progression. However, as most of the evidence was gathered in the cytokine era, the importance of CRP should be re-evaluated as a useful biomarker for RCC in today's molecular-targeted era.

Prognostic value of CRP in urothelial carcinoma

The significance of CRP in the prediction of outcome has also been investigated in UC of the UUT and bladder in the same context as RCC. The prognostic impact of CRP has also been shown in this field (Table 2).

Table 2. Prognostic significance of CRP in UC
StudyPatientsDiseaseTreatmentCut-off of CRP (mg/L)Prognostic value of CRP
Himly et al.68105BladderResection10CRP is significantly associated with CSS.
Saito et al.69130UUT;Surgery5CRP is a significant prognostic factor for RFS and CSS.
T: any; N: any, M: 0
Yoshida et al.7088MIBCChemoraditotherapy5CRP predicts CSS. Non-normalization of CRP after chemoradiotherapy indicated poor prognosis.
Gakis et al.71246BladderCystectomy5CRP is combined into a new prognostic model of TNR-C to predict CSS.
Ishioka et al.72223Locally advanced or metastaticMultimodal treatment• (Continuous)CRP improves the predictive accuracy of the nomogram for the estimation of survival probability.
Saito et al.7380Locally advanced or metastaticChemotherapy5CRP kinetic predicts OS. Normalized CRP period could directly predict overall survival period.

In our first report on UC with 130 UUT-UC patients treated surgically, preoperative CRP level was found to be an independent prognostic factor for RFS and CSS.69 Among the patients followed for a median of 47 months, elevated preoperative CRP levels (5 mg/L or greater) in 24 patients were associated with poor survival, with 5-year RFS and CSS rates of 21 and 25%, respectively, in the elevated CRP group; and 59 and 80%, respectively, in the non-elevated CRP group. HR of elevated CRP on CSS was 1.78 compared with those of pT3 or greater, and lymph node involvement of 1.77 and 2.25, respectively.

In a study by Hilmy et al. of 105 patients with bladder cancer who underwent resection, CRP was found to be significantly associated with survival as well as tumor stage.68 Among 59 patients whose preoperative CRP levels were available, the mean CSS period of 65.5 months in patients with preoperative CRP levels of >10 mg/L was shorter than that of 103.7 months in patients with normal CRP levels. We previously reported that elevated pretreatment CRP concentration (5 mg/L or greater, HR 1.80) was associated with poor CSS in 88 patients with muscle-invasive bladder cancer treated with chemoradiotherapy as well as tumor stage (pT3 or grater, HR 1.87).70 The CSS of the elevated CRP group was significantly lower than that of the non-elevated CRP group. Furthermore, in this study, we reported that peritreatment CRP kinetics had additional impact on survival. Among patients with elevated pretreatment CRP levels, those whose post-treatment CRP levels remained elevated had extremely poor prognosis compared with those whose post-treatment CRP levels normalized at less than 5 mg/L.

Based on the significance of CRP in predicting the prognosis for UC, a new prognostic algorithm incorporating CRP levels for invasive bladder cancer has been proposed, as demonstrated in RCC. The prognostic model, TNR-C score, was developed from 246 patients with bladder cancer who underwent radical cystectomy.71 TNR-C was composed of four elements: tumor stage (T), lymph node density (N) and resection margin status (R), and the presence or absence of elevated CRP levels (C). The score was 4 for positive resection margins, 3 for tumor stage greater than pT3a, 2 for lymph node density ≥0.09, and 1 for CRP concentration of >5 mg/L or 0 (without those features), respectively. The sum of these scores ranged between 0 and 10; high scores indicated poor prognosis with 5-year CSS rates of 78.1, 42.8 and 0% for patients with scores 0–2, 3–6 and 7–10, respectively. CRP improved the predictive accuracy by 4.9% with a concordance index of 0.788 in their final prognostic model. We have shown the prognostic impact of CRP in patients with locally advanced or metastatic UC, and developed a nomogram incorporating CRP for survival probability, in which CRP improved the predictive accuracy.72 Those studies showed that CRP serves as a prognostic factor for both locally confined and metastatic UC in prognostic algorithms.

As shown in RCC, we previously reported that CRP kinetics obtained by serial measurement of CRP levels is also useful for the evaluation of disease outcome in UC patients.73 In the study, 80 patients receiving cisplatin-based systemic chemotherapy as a second-line setting were divided into three groups (non-elevated, normalized and non-normalized CRP group) according to their baseline and nadir CRP levels, using a cut-off value of 5 mg/L as in our previous report on RCC.56 CRP kinetics status had a significant impact on survival with 1-year OS rates of 72, 51 and 14%, and median survival periods of 23, 13 and 6 months for non-elevated, normalized and non-normalized CRP groups, respectively. Furthermore, normalized CRP periods were also significantly associated with OS periods in those patients. The current evidence indicates that CRP can serve as a biomarker that satisfies the NIH criteria for UC.

Prognostic value of CRP in prostate cancer

In prostate cancer, PSA has served as a definitive biomarker to estimate the probability and extent of cancer, to monitor the response to therapeutic intervention, and to predict outcome in clinical practice. Along with PSA, CRP has been shown to be prognostic in some clinical settings, such as advanced disease and castration-resistant cancer (Table 3).

Table 3. Prognostic significance of CRP in prostate cancer
StudyPatientsStageTreatmentCut off of CRP (mg/L)Prognostic value of CRP
McArdle et al.7498LocalizedRadical treatment3, 10CRP is significantly associated with CSS and OS.
Nakashima et al.75126Metastatic (bone)Endocrine therapy1.5CRP is a significant prognostic factor for CSS.
McArdle et al.7662MetastaticEndocrine therapy10CRP is an independent prognostic factor for CSS.
Beer et al.77160CRPCPhase II trial (including docetaxel)8CRP is an independent prognostic factor for OS.
Peer et al.78119CRPCPhase II trial (including docetaxel)5CRP is associated with OS.
Ito et al.7980CRPCDocetaxel5CRP is a significant prognostic factor for OS.

CRP concentrations were more highly correlated with the extension of bone metastasis in patients with prostate cancer who had bone metastasis than those without metastasis.80 However, in a large population-based study, no association or trend was observed between CRP level and risk of prostate cancer.81 As a positive association between high-sensitivity CRP tests that can measure very low levels of CRP (<1 mg/L) and the risk of cancer has been reported,82 further studies on the influence of the inflammatory response that can be detected by such a sensitive assay on prostate cancer are warranted.

In a study by McArdle on 98 localized prostate cancer patients, CRP was found to be an independent prognostic factor for patients, as was age and Gleason score.74 The elevated CRP levels found in 38 patients (>3 mg/L) were associated with poor CSS and OS (HR 1.88 and 1.60, respectively) as well as age (>70 years, HR 4.88 and 4.43, respectively) and Gleason score (HR 2.16 and 1.57, respectively).

The importance of CRP as a prognostic factor is emphasized, as it relates to the situation of advanced disease in prostate cancer. Nakashima et al. showed that CRP, as well as the extent of disease on bone scan, is an independent prognostic factor for prostate cancer with bone metastasis.75 Elevation of CRP level (>1.5 mg/L) was associated with poor survival in 126 patients (HR 1.88), as well as the extent of disease of bone metastasis (extent of disease 2 or greater, HR 2.24). In another study of 62 patients with metastatic prostate cancer, CRP (cut-off point: 10 mg/L, HR 1.97) and PSA (HR 1.96) were independent prognostic factors for CSS.76

In the more advanced state of CRPC, CRP might become more significant as a prognosis predictor, because the prognostic power of PSA might be reduced. Two phase II clinical trials that include docetaxel in the protocols reported the significance of CRP as a prognostic indicator. In the Androgen-Independent Prostate Cancer Study of Calcitriol Enhancing Taxotere trial, patients with metastatic CRPC were given docetaxel weekly at a dose of 36 mg/m2 with calcitriol (45 µg) or placebo on the day before docetaxel treatment, for three consecutive weeks in a 4-week cycle.77 Among 160 patients whose pretreatment CRP levels were available, CRP levels were elevated in 102 patients and CRP was an independent prognosticator for OS as both continuous (HR 1.41) and categorical (cut-off point: 8 mg/L, HR 2.96) variables. The other trial in 119 patients with CRPC showed that elevated CRP levels of 5 mg/L or greater were associated with poor OS (HR 1.11).78

We also demonstrated that CRP was an independent prognostic factor for CRPC treated with docetaxel (cut-off point 5 mg/L, HR 1.95), as well as hemoglobin concentration (cut-off point 11 g/dL, HR 1.63).79 In a practical cohort of 80 patients, 34 patients showed elevated CRP levels of 5 mg/L or greater, and CRP had a significant impact on OS. The median survival periods of patients with elevated CRP levels were 25 months compared with 11 months in those without elevated CRP. Taken together, these findings indicate that CRP also has the potential to be a biomarker for prostate cancer.

CRP as a prognosticator for late complications in testicular cancer

As for prostate cancer, it is also well known that testicular cancer has definitive biomarkers, such as lactate dehydrogenase, alpha-fetoprotein and human chorionic gonadotropin, all of which are used to evaluate the risk of mortality, to monitor the efficacy of therapeutic intervention, and to determine treatment strategies in clinical practice. CRP has not been shown to have a prognostic impact in terms of predicting disease-specific mortality resulting from the testicular cancer itself; however, it could play a role in predicting late complications.

The risk of cardiovascular disease and second non-germ cell malignancies increased in long-term survivors of testicular cancer who received systemic therapy.83–85 As CRP is recognized as a significant biomarker for various disorders, including cardiovascular disease35 and malignancies, CRP acts as a predictor for the subsequent development of these disorders.

Among long-term survivors of testicular cancer, elevated CRP levels were found more often in patients who had received radiotherapy (1.5 mg/L) compared with those who had only received chemotherapy or surgery (1.2 mg/L for both).86 In one study of patients with testicular cancer treated between 1980 and 1994, patients completed the first survey between 1998 and 2001; the survey included a measurement of CRP level. The second survey was carried out between 2007 and 2008. A significant association between elevated CRP level and increased risk of subsequent cardiovascular disease (HR 2.79), and subsequent second non-germ cell cancer (HR 2.21) was observed in comparison with non-elevated CRP level.87

High concentrations of CRP were linked to chronic cancer-related fatigue in testicular cancer survivors, as well as with circulating levels of IL-1 receptor antagonist.88 This association might be mediated by physiological rather than psychological conditions, as only the physical elements, and not the psychological elements, correlated with CRP in the fatigue questionnaire. These findings suggest that a low-grade inflammatory process might cause chronic cancer-related disorders in testicular cancer survivors. CRP might serve as a biomarker for chronic cancer-related disorders in the long-term surveillance of survivors.

Conclusions

CRP, a representative marker for systemic inflammatory response, has a significant impact in predicting outcomes of urological cancer. The current evidence that suggests that inflammation is involved in the pathogenesis of cancer progression supports the strong association between the presence of systemic inflammatory response as evidenced by elevated CRP level and poor outcome in urological cancers.

Based on its strong prognostic power, CRP can realize simpler prognostic algorithms that retain their predictive abilities with established parameters, as demonstrated for RCC and UC. By including CRP in prognostic algorithms, the estimation of individual risk and the stratification of patients for clinical trials are made easier.

Serial measurement of CRP is also useful for monitoring the disease course, such as the effect of treatment intervention or further progression. For example, CRP kinetics, presented with a stratification model of longitudinal changes in CRP levels according to baseline and nadir CRP levels, could add further information that would enable an evaluation of the clinical course of urological cancer. Further application of serial CRP level changes, such as their incorporation into prognostic algorithms and decision trees, is warranted as it would enable clinicians to assess CRP levels and would be a useful tool in the management of urological cancers.

These findings allow us to conclude that CRP might serve as a useful biomarker for urological cancers and that it satisfies the 2001 NIH criteria. CRP is already measured objectively and affordably in clinical practice worldwide. CRP can be used as an indicator of disease outcome and to monitor the disease course. A better understanding of the association between the presence of systemic inflammatory response and cancer progression might provide insights into novel cancer therapeutics. Anti-inflammatory therapeutics that target the tumor microenvironment might also be considered in the future.

Conflict of interest

None declared.

Ancillary