SEARCH

SEARCH BY CITATION

Keywords:

  • renal cell carcinoma;
  • regulatory T cells;
  • cyclooxygenase-2;
  • tumour-infiltrating lymphocytes;
  • prognosis

Abstract

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

OBJECTIVE

To investigate the prognostic value of regulatory T cells (Tregs) and its correlation with cyclooxygenase-2 (COX-2) expression in clear cell renal cell carcinoma (RCC).

PATIENTS AND METHODS

CD4+, Foxp3+ tumour-infiltrating lymphocytes and tumour COX-2 expression were assessed by immunohistochemistry in tissue microarrays containing RCC from 125 patients. Prognostic effects of low and high expression were evaluated by Cox regression and Kaplan–Meier analysis using the median values as thresholds. The expression of Tregs and COX-2 were compared with the clinicopathological variables. In addition, Tregs and its correlation with COX-2 expression was also analysed.

RESULTS

Peritumoral Tregs were positively correlated with intratumoral COX-2 expression (Spearman rank correlation 0.336, P < 0.001). Peritumoral Tregs were associated with TNM stage (P = 0.001) and tumour size (P = 0.002), while intratumoral COX-2 expression was associated with TNM stage (P = 0.018) and grade (P = 0.013). Using multivariate analysis, increased peritumoral Tregs, higher TNM stage (III + IV), larger tumour size (≥7 cm) and higher nuclear grade (III + IV) were independent predictors for significantly shorter overall survival and disease-free survival.

CONCLUSIONS

Increased peritumoral Tregs are associated with worse prognosis in clear cell RCC. The high intratumoral COX-2 expression may be the underlying reason for the aberrant gathering of Tregs. These results suggest that clinical application of COX-2 inhibitors may benefit those patients with higher intratumoral COX-2 immunostaining by reducing the transformation of Tregs in RCC.


Abbreviations
TILs

tumour-infiltrating lymphocytes

Tregs

regulatory T cells

OS

overall survival

DFS

disease-free survival

COX-2

cyclooxygenase-2

PGE2

prostaglandin E2.

INTRODUCTION

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

It has long been known that RCC is relatively insensitive to cytotoxic agents and radiotherapy, and the most promising agents used in the treatment of RCC are biological response modifiers such as interleukin-2 and interferon-α[1]. In addition, there are some reports verifying spontaneous regression of pulmonary metastases from RCC after the ablation of the primary tumour [2]. These reports indicate the important role of the immune system in the formation and progression of RCC. Indeed, pathological specimens of RCC frequently harbour large numbers of tumour-infiltrating lymphocytes (TILs) [3–5]. TILs are considered manifestations of host immune reactions against cancers. Patients that underwent resection for RCC with a prominent lymphocytes infiltrate, especially T lymphocytes, had increased recurrence and shorter survival [6], which is different from other malignancies [7]. So, are there some unknown regulatory factors in the local environment of RCC that hamper effective immune responses? CD4+CD25+ regulatory T cells (Tregs) are thought to dampen T-cell immunity and to be the main obstacle tempering immunotherapy [8]. Foxp3, a critical regulator of Tregs’ development and function, fulfils the criteria of a Treg-specific marker [9–11]. High Foxp3 expression is not only associated with poor prognosis in ovarian cancer but also represents an independent predictor for overall survival (OS) and progression-free survival [12]. However, discrepancies remain among different cancer types. In RCC, the prognostic influence of Tregs are contradictory. Siddiqui et al.[13] reported that intratumoral Tregs had no prognostic value in RCC, at the same time peritumoral Tregs were omitted in their experiment. Griffiths et al. [14] reported an association between higher peripheral blood Tregs count and adverse OS. However, the mechanisms that underlie the aberrant enhancement of Tregs in RCC are not elucidated in these investigations.

Cyclooxygenase (COX) is a rate-limiting enzyme involved in the conversion of arachidonic acid to prostaglandin E2 (PGE2). Two COX genes, COX-1 and COX-2, have been identified. COX-1 is constitutively expressed in many tissues and is involved in several physiological functions including the cytoprotection of the stomach, vasodilatation in the kidney, and the production of a proaggregatory prostanoid, thromboxane by the platelets. On the other hand, COX-2 is an inducible gene originally found to be induced by inflammation or by various other stimuli, such as mitogens, cytokines, and growth factors. Previous studies have stressed the potential role of COX-2 in carcinogenesis and the induction of COX-2 has been reported in colorectal, gastric, breast, oesophagus, and lung carcinomas. Similarly, immunohistochemistry results show that COX-2 is also highly expressed in RCC [15,16]. Recently, a report on lung cell carcinoma showed the important role of COX-2-derived PGE2 in the transformation of Tregs [17]. In view of this, we hypothesize that COX-2 in RCC may play an important role in the development of cancer through transformation of Tregs, in addition to other known mechanisms, such as induction of angiogenesis and suppression of apoptosis.

To our knowledge, research on the immunohistochemical expression of Tregs and its relationship with COX-2 has not been performed in clinical tissues of RCC. In the present study, we investigated the prognostic value of both intratumoral and peritumoral Tregs in RCC, in addition to elucidating whether there was a correlation between Tregs and COX-2 expression.

PATIENTS AND METHODS

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

We identified 125 patients with RCC treated with radical nephrectomy or nephron-sparing surgery for unilateral, sporadic, clear cell RCC at Zhong Shan Hospital (Shanghai, China) between 1999 and 2005 after giving informed consent. None of these patients received preoperative immunotherapy, renal arterial embolization therapy or COX-2 inhibitors chronically. Paraffin blocks were selected only based on the availability of suitable paraffin-embedded tissue with complete clinicopathological and follow-up data for the patients. Ethical approval was obtained from the Zhong Shan Hospital Research Ethics Committee. The patients’ clinicopathological characteristics are summarized in Table 1.

Table 1.  Pathological features of the 125 patients with clear cell RCC
CharacteristicsN (%) of patients
Gender 
 Male79 (63.2)
 Female46 (36.8)
TNM stage groupings 
 I90 (72)
 II25 (20)
 III 3 (2.4)
 IV 7 (5.6)
Primary tumour size, cm 
 <564 (51.2)
 5–730 (24)
 7–1021 (16.8)
 >1010 (8)
Nuclear grade 
 117 (13.6)
 289 (71.2)
 315 (12)
 4 4 (3.2)

The follow-up was finished on 24 October 2007. The duration of follow-up was calculated from the date of nephrectomy to the date of cancer progression (i.e. distant metastases after nephrectomy for the primary tumour), death, or last follow-up. The median (range) follow-up was 57.2 (5.5–102) months. All patients were prospectively monitored by serum creatinine, abdomen ultrasonography and chest X-ray every 1–6 months according to the postoperative time and individual factors. For suspicious cases, CT and/or MRI were used to verify whether metastases had developed. The treatment method after relapse varied, 21 metastatic patients were treated with interferon-α or combined therapies, another five patients received no treatment because of health problems or economic reasons, finally there was no complete response in any of these patients. OS was defined as the interval between surgery and death or between surgery and the last observation for surviving patients. The data were censored at the last follow-up for living patients.

The pathological features examined included tumour size, primary tumour classification (T), regional lymph node involvement (N), distant metastases at nephrectomy (M), and nuclear grade. The tumours were staged according to the 2002 TNM staging system [18]. Nuclear grade was assigned according to the criteria of Fuhrman et al. [19]. One pathologist (J.H) reviewed all specimens with no knowledge of the patient outcome.

Tissue microarrays were constructed as previously described [20–23]. All RCC cases were histologically reviewed by haematoxylin and eosin staining, and representative areas with small round lymphocyte infiltrate were premarked in the paraffin blocks, away from necrotic and haemorrhagic materials. Duplicate 1-mm diameter cylinders from two different areas, tumour centre and nearest noncancerous margin (designated as intratumour and peritumour, respectively), that is, a total of four punches, were included in each case, together with different controls, to ensure reproducibility and homogenous staining of the slides (Shanghai Biochip Co., Ltd, Shanghai, China). Thus, four different tissue microarray blocks were constructed, each containing 125 cylinders. Sections of 4-µm thickness were taken on 3-aminopropyltriethoxysilane-coated slides. The mouse monoclonal antibodies used were: anti-human CD4 (Cayman), Foxp3 (Biolegend), COX-2 (Cayman). Immunohistochemistry of paraffin sections was carried out using a two-step protocol (Novolink Polymer Detection System, Novocastra, Newcastle, UK) according to the manufacturer’s instructions. Briefly, paraffin sections were first deparaffinized and then hydrated. After microwave antigen retrieval as required, endogenous peroxidase activity was blocked with incubation of the slides in 0.3% H2O2 and nonspecific binding sites were blocked with Protein Block (RE7102). After incubation with primary antibodies, Post Primary Block (RE7111) and secondary antibody (Novolink Polymer RE7112) serially, the sections were developed in 3,3’-diaminobenzidine solution under the microscope and counterstained with haematoxylin. The control tissue used for immunostaining of CD4 and Foxp3 were lymph nodes, while control tissue for immunostaining of COX-2 was a breast tumour with known over-expression of COX-2 enzyme by immunohistochemistry. Negative control slides omitting the primary antibodies were included in all assays.

To evaluate immunohistochemical variables, the number of CD4+ and Foxp3+ T cells was counted as previously described [23]. Briefly, we used a computerized image analysis system composed of a camera (HV-C20A CCD, Hitachi, Tokyo, Japan), installed on a Leica DMLA light microscope (Leica Microsystems, Wetzlar, Germany) and attached to a personal computer. Under × 400 magnification, there existed at least 12 independent and intact computerized microscopic fields for the duplicate of each case. We counted every lymphocytic infiltrates for each patient sample to ensure representativeness and homogeneity. The results were expressed as the mean (se) number cells for one case. For COX-2 expression, the weighted score was computed that represented the product of the percentage of tumour cell positivity and intensity, as described previously [24]. Briefly, under × 200 magnification, the percentage tumour cell positivity was categorized as follows: 0, <5%; 1, 5–25%; 2, 26–50%; 3, 51–75%, and 4, >75%. Staining intensity was graded as absent (0), weak (1), medium (2), or strong (3). The evaluation of the above indexes was performed without knowledge of the clinicopathological data by two investigators. Variations in the enumeration, within a range of 5%, were re-evaluated and a consensus decision was made.

Cumulative survival time was calculated by the Kaplan–Meier method and analysed by the log-rank test. Univariate and multivariate analyses were based on the Cox proportional hazards regression model. For all the immunohistochemical markers, the threshold for definition of subgroups was the median value. A secondary analysis was performed to assess the relationship among two immunohistochemical variables and clinicopathological characteristics. Chi-squared and Fisher’s exact tests were used as appropriate for comparing individual variables. In addition, the correlation between intratumoral COX-2 and peritumoral Tregs were evaluated using Spearman rank correlation. A two-tailed P < 0.05 was considered to indicate statistical significance.

RESULTS

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

IMMUNOHISTOCHEMICAL CHARACTERISTICS

Using haematoxylin and eosin staining, immune cell infiltration varied substantially among different patients. Lymphocytes infiltrated RCC in a diffuse manner or in lymphoid aggregates, with more abundant cells in peritumoral areas. The duplicate of spots for each tumour showed a good level of homogeneity for stained cell density. Cells were considered positive for CD4+ based on the presence of circular membranous staining. Cells were considered positive for Foxp3+ based on distinct intranuclear expression (Fig. 1a,b). Tumour cells were considered positive for COX-2 based on cytoplasmic or membranous staining (Fig. 1c,d). The statistics for the immunohistochemical variables are shown in Table 2.

image

Figure 1. Immunohistochemical staining for peritumoral Foxp3+ (a, b; × 400) and intratumoral COX-2 (c, d; × 200); positive cells were stained brown. a, High peritumoral Foxp3+ expression ). b, Almost no peritumoral Foxp3+. c, Almost all cancer cell cytoplasm is strongly stained for COX-2 (COX-2 score 12). d, Almost all cancer cells are not stained (COX-2 score 0).

Download figure to PowerPoint

Table 2.  Descriptive statistics of immunohistochemical variables
VariableMean (se)Median (range)
CD4  
 Intratumoral26.89 (26.94)21.00 (1.00–146.00)
 Peritumoral54.52 (49.87)40.50 (2.00–234.00)
Foxp3  
 Intratumoral2.21 (4.02)0.50 (0–28.50)
 Peritumoral5.52 (7.86)2.50 (0–48.50)
COX-2  
 Intratumoral5.70 (3.63)6.00 (0–12.00)
 Peritumoral2.00 (2.43)1.00 (0–9.00)

PROGNOSTIC FACTORS

At last follow-up, 27 of the 125 patients studied had died, including 26 patients who died from RCC (one patient died from heart disease). The OS and disease-free survival (DFS) rates were 92% and 82% at 1 year, 81% and 80% at 3 years, 78% and 78% at 5 years, and 68% and 71% at 7 years, respectively, for the whole study population.

On univariate analysis, age and sex had no prognostic significance for OS and DFS. Also, intratumoral and peritumoral CD4+ TILs were not associated with OS or DFS. TNM stage, tumour size and nuclear grade were predictors for both OS and DFS (Table 3). High peritumoral Tregs (Foxp3+) were prognostic for both worse OS (P < 0.001) and DFS (P < 0.001) (Table 3; Figs 2a,3a). High intratumoral COX-2 was also considered to be prognostic for both reduced OS (P = 0.021) and DFS (P = 0.027) (Table 3; Figs 2b,3b). In addition, both peritumoral COX-2 and intratumoral Treg cells had no prognostic value in RCC (Table 3).

Table 3.  Univariate analyses of factors associated with OS and DFS
VariableOSDFS
HR (95%CI)PHR (95%CI)P
  1. HR, hazard ratio.

Age (<65 vs ≥65 years)0.871 (0.389–1.949)0.7360.873 (0.390–1.951)0.740
Sex (male vs female)0.776 (0.337–1.787)0.5510.733 (0.319–1.686)0.465
Tumour size (<7 vs ≥7 cm)6.858 (3.075–15.293)<0.0015.456 (2.469–12.059)<0.001
Nuclear grade (I–II vs III–IV)3.924 (1.792–8.595)0.0013.697 (1.688–8.097)0.001
TNM stage (I–II vs III–IV) 11.259 (4.657–27.221)<0.0017.926 (3.326–18.885)<0.001
Peritumoral factors (low vs high)    
 CD4+1.135 (0.530–2.431)0.7441.078 (0.504–2.306)0.847
 Foxp3+6.878 (2.357–20.069)<0.0016.738 (2.309–19.666)<0.001
 COX-21.821 (0.818–4.055)0.1421.737 (0.780–3.867)0.176
Intratumoral factors (low vs high)    
 CD4+1.481 (0.677–3.239)0.3251.455 (0.668–3.171)0.345
 Foxp3+1.384 (0.644–2.974)0.4051.374 (0.642–2.941)0.413
 COX-22.568(1.115–5.915)0.0212.473(1.073–5.697)0.027
image

Figure 2. Kaplan–Meier overall survival curve according to (a) peritumoral Foxp3+ expression (high, green; low, blue), where the survival of patients with a high peritumoral Foxp3+ expression was significantly worse than that of patients with a low peritumoral Foxp3+ expression (P < 0.001); (b) intratumoral COX-2 expression (high, green; low, blue), where the survival of patients with a high COX-2 expresson was significantly worse than that of patients with a low COX-2 expression (P = 0.021).

Download figure to PowerPoint

image

Figure 3. Kaplan–Meier DFS curve according to (a) peritumoral Foxp3+ expression (high, green; low, blue), where the survival of patients with a high peritumoral Foxp3+ expresson was significantly worse than that of patients with a low peritumoral Foxp3+ expression (P < 0.001); (b) intratumoral COX-2 expression (high, green; low, blue), where the survival of patients with a high COX-2 expresson was significantly worse than that of patients with a low COX-2 expression (P = 0.027).

Download figure to PowerPoint

The clinicopathological features showing significance by univariate analysis were further adopted as covariates when performing the multivariate Cox proportional hazards model. Increased peritumoral Tregs, higher TNM stage (III + IV), larger tumour size (≥7 cm) and higher nuclear grade (III + IV) were independent predictors for significantly shorter DFS and OS (Table 4). However, intratumoral COX-2 was no longer an independent prognostic factor (Table 4).

Table 4.  Multivariate analyses of factors associated with OS and DFS
VariableOSDFS
HR (95%CI)PHR (95%CI)P
  1. HR, hazard ratio.

Tumour size (<7 vs ≥7 cm)3.269 (1.365–7.827)0.0082.496 (1.047–5.948)0.039
Nuclear grade (I–II vs III–IV)3.806 (1.501–9.655)0.0052.766 (1.169–6.541)0.021
TNM stage (I–II vs III–IV)7.899 (2.728–22.872)<0.0013.720 (1.428–9.694)0.007
Peritumoral Foxp3+ (low vs high)4.264 (1.291–14.084)0.0174.114 (1.262–13.414)0.019
Intratumoral COX-2 (low vs high)0.641 (0.223–1.838)0.4080.816 (0.300–2.220)0.691

Peritumoral Tregs were associated with TNM stage (P = 0.001) and tumour size (P = 0.002), while intratumoral COX-2 was associated with TNM stage (P = 0.018) and nuclear grade (P = 0.013) (Table 5). Furthermore, peritumoral Tregs were correlated with intratumoral COX-2 expression (Spearman rank correlation r = 0.336, P < 0.001).

Table 5.  Relationship between intratumoral COX-2, peritumoral Foxp3+ expression and clinicopathological characteristics
VariableCOX-2Foxp3+
LowHighPLowHighP
  • Fisher’s exact tests, and chi-squared tests for all the other analyses.

Patients, n6065 6065 
Age, years      
 <6537450.45237450.452
 ≥652320 2320 
Sex      
 Male36430.57837420.853
 Female2422 2323 
Tumour size, cm      
 <748460.30153410.002
 ≥71219  724 
Nuclear grade      
 I–II56500.01354520.140
 III–IV 415  613 
TNM stage      
 I–II59560.01860550.001
 III–IV 1 9  010 

DISCUSSION

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

Although developments in diagnostic imaging techniques have enabled the early detection of RCC [25,26], 20–30% of patients with newly diagnosed RCC have evidence of metastases at presentation. In addition, another 25–30% of patients treated for localized RCC will experience metastatic disease progression after surgical extirpation [27,28]. The median survival for patients with metastatic RCC is 6–10 months with a 20% survival rate at 2 years [29]. The lack of effective systemic therapy for metastatic RCC is, in part, due to a fundamental lack of understanding of the molecular events that result in cellular transformation, carcinogenesis and tumour progression in the human kidney.

Tregs have a crucial role in impeding immune surveillance against cancer and hampering the development of effective antitumour immunity [30]. It has been reported that the frequencies of Tregs are increased in both peripheral blood and tumour microenvironment in RCC, and Tregs from RCC can suppress proliferation of autologous T cells in vitro[13]. One clinical experiment in a series of patients with metastatic RCC indicated that elimination of Tregs is capable of enhancing the immunostimulatory efficacy of tumour RNA-transfected dendritic cell vaccines [31]. All these experiments indicate the important role of Tregs in RCC. However, one investigation reported that tumour infiltrating Foxp3+CD4+CD25+ T cells had no prognostic value in RCC [13].

In the present study, we showed that high Tregs in the peritumoral areas predicted poor survival in RCC, while intratumoral Tregs had no prognostic value in RCC (Table 3). Correspondingly, intratumoral Tregs also had no prognostic value in the Siddiqui et al. experiment [13]. Several aspects may contribute to the present results: primarily, we assessed both intratumoral and peritumoral areas, while only the intratumoral areas were assessed in the Siddiqui et al. experiment. In the present study, there were fewer Tregs present in the intratumoral areas of RCC compared with peritumoral areas (2.21 vs 5.52). Previous research [6] has identified the elevated presence of mononuclear cells in peritumoral areas. In addition, in the present study peritumoral Tregs were positively associated with TNM stage and tumour size, which have long been known to be prognostic factors in RCC. Secondly, in the present study, we used tissue microarray techniques, which have been shown to be a powerful tool for evaluating tumours simultaneously with histological and immunohistochemical analyses [32].

In the present study peritumoral Tregs were positively correlated with intratumoral COX-2 expression (P < 0.001). In addition, a previous report in lung cell carcinoma showed the important role of COX-2 derived prostaglandins (PGE2) in the transformation of Tregs [17]. Therefore, the overexpression of COX-2 in RCC could be the underlying reason for the aberrant gathering of Tregs. COX-2 inhibition may dampen the transformation of Tregs and in turn contribute to eradicating RCC.

In summary, the present results indicate that increased peritumoral Tregs is associated with worse prognosis in clear cell RCC. High intratumoral COX-2 expression may be the underlying reason for the aberrant gathering of Tregs in RCC. Therefore, COX-2 inhibition may dampen the transformation of Tregs and in turn contribute to eradicating RCC. The clinical application of a COX-2 inhibitor may benefit those patients with higher intratumoral COX-2 immunostaining.

ACKNOWLEDGEMENTS

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

The financial support of the Basic and Clinical Medicine cross-disciplinary Research Foundation of Shanghai Medical College of Fudan University (Grant no. 2007 JL01) and Scientific research start-up foundation for Young major scientists of Shanghai Medical College of Fudan University (Grant no. 2007 L02) are gratefully acknowledged. We also thank Shanghai Biochip Co., Ltd. for help in tissue microarrays construction.

REFERENCES

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