ResearchEvaluation of neuroendocrine markers in renal cell carcinoma

Background: The purpose of the study was to examine serotonin, CD56, neurone-specific enolase (NSE), chromogranin A and synaptophysin by immunohistochemistry in renal cell carcinomas (RCCs) with special emphasis on patient outcome. Methods: We studied 152 patients with primary RCCs who underwent surgery for the removal of kidney tumours between 1990 and 1999. The mean follow-up was 90 months. The expression of neuroendocrine (NE) markers was determined by immunohistochemical staining using commercially available monoclonal antibodies. Results were correlated with patient age, clinical stage, Fuhrman grade and patient outcome. Results: Eight percent of tumours were positive for serotonin, 18% for CD56 and 48% for NSE. Chromogranin A immunostaining was negative and only 1% of the tumours were synaptophysin immunopositive. The NSE immunopositivity was more common in clear cell RCCs than in other subtypes (p = 0.01). The other NE markers did not show any association with the histological subtype. Tumours with an immunopositivity for serotonin had a longer RCCspecific survival and tumours with an immunopositivity for CD56 and NSE had a shorter RCC-specific survival but the difference was not significant. There was no relationship between stage or Fuhrman grade and immunoreactivity for serotonin, CD56 and NSE. Conclusions: Serotonin, CD56 and NSE but not synaptophysin and chromogranin A are expressed in RCCs. However, the prognostic potential of these markers remains obscure. Background Neuroendocrine (NE) cells are important for regulating cell growth and differentiation. In addition to specific NE tumours, NE activity can be detected in other types of tumours such as breast [1] or prostate carcinomas [2]. The specific NE tumours of kidney include carcinoid, NE carcinoma, primitive neuroectodermal tumour, neuroblastoma and phaeochromocytoma [3]. NE tumours can show a wide range of behaviour. Small cell carcinomas of the lung are aggressive [4], whereas carcinoid tumours show indolent behaviour [5]. In patients with prostate adenocarcinoma, NE differentiation has been linked to both aggressive behaviour [6] and better survival [7]. Serotonin (5-hydroxytryptamine, 5HT) is a growth factor for several types of malignant cells. Serotonin causes cellular proliferation [8], and there is also evidence linking it to oncogenes [9]. By contrast, serotonin can also inhibit tumour growth because of its vasoconstrictive effect [10]. In RCC patients, plasma levels of serotonin [11] and the immunoexpression for serotonin have previously been examined in patients with advanced disease [12]. As far as we know, the prognostic significance of serotonin expression in RCC patients has not been studied in large RCC patient material. CD56 is neural cell adhesion molecule (NCAM), which is also found in some lymphocytes [13]. In terms of clinical pathology, CD56 is a rather sensitive indicator of NE differentiation. The immunoexpression of CD56 has previously been studied in RCC and its prognostic potential in the survival of RCC patients has also been evaluated [14]. Neurone-specific enolase (NSE) is a broad-spectrum, non-specific NE marker of all types of neurons, NE or paraneuronal cells and even various malignant tumours of non-NE types [15]. Its serum levels and immunoexpression have also been studied in RCCs [16,17,12]. * Correspondence: hanna.ronkainen@mail.suomi.net 1 Department of Surgery, PO Box 21, Oulu University Hospital, FIN-90029 Oulu, Finland Full list of author information is available at the end of the article BioMed Central © 2010 Ronkainen et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Ronkainen et al. Diagnostic Pathology 2010, 5:28 http://www.diagnosticpathology.org/content/5/1/28 Page 2 of 7 Chromogranin A is an abundant monomeric protein in the neurosecretory granules of NE cells, and its immunostaining correlates to the number of NE granules seen at the level of electron microscopy [13]. The serum levels and immunoexpression of chromogranin A have previously been studied in RCCs to evaluate its prognostic significance [12,17]. Synaptophysin is regarded as one of the basic markers of NE differentiation. It is an integral part of the NE secretory granule membrane [13]. To our knowledge, the immunoexpression of synaptophysin has not previously been studied in RCCs. The aim of this study was to clarify the extent of the immunoexpression of NE markers, serotonin, CD56, NSE, chromogranin A and synaptophysin in RCCs and their significance regarding the behaviour of these tumours. For this we investigated a large set of RCCs consisting of different histological types and correlated the results with the clinical behaviour of the tumours. Methods The retrospective study group consisted of 152 patients treated with radical nephrectomy or renal resection for primary RCC at Oulu University Hospital, Oulu, Finland between 1990 and 1999. Patients underwent medical examination and preoperative staging including chest Xray and/or thoracic CT and abdominal CT. The research plan was approved by the local ethics board. All the data from the patients' records and Finnish Cancer Registry were re-evaluated by the same urologist. The exact stage of the disease was recorded according to the TNM classification of RCCs [18]. Archival material of formalin-fixed and paraffinembedded tumours were reclassified and graded according to current WHO classification [3]. The most representative area from each tumour block was selected to a multitissue array block. The array section was 3 μm thick. Immunostaining procedure The antibodies used in the immunostaining were monoclonal mouse anti-human serotonin (DakoCytomation, Glostrup, Denmark) in a dilution of 1:200, lyophilized mouse monoclonal antibody for CD56 (Novocastra Laboratories Ltd., Newcastle-upon-Tyne, UK) in a dilution of 1:200, monoclonal mouse anti-NSE (Zymed Laboratories, Carlsbad, CA, USA) in a dilution 1:1000, polyclonal rabbit anti-chromogranin A (Zymed Laboratories, Carlsbad, CA, USA) in a dilution 1:500 and monoclonal mouse anti-synaptophysin (DakoCytomation) in a dilution 1:50. First, the sections were deparaffinised in xylene, rehydrated in descending ethanol series and washed in phosphate-buffered saline (PBS). Then, the sections were boiled in 0.01 M citrate buffer (pH 6) for 10 min (serotonin and CD56) or Tris/EDTA for 15 min (chromogranin A and synaptophysin) in a microwave oven. The sections were cooled for 15 min and washed twice in PBS. Endogenous peroxidise activity was eliminated by incubation in 5% hydrogen peroxide and absolute methanol. Bound antibodies were visualised using an UltraVision (Thermo Fisher Scientific, Fremont, CA, USA) for serotonin, CD56 and NSE and EnVision (DakoCytomation) for chromogranin A and synaptophysin. DAB (5% 3,3'diaminobenzidine tetrahydrochloride, DakoCytomation) was used as the chromogen. The positive controls for stainings were the metastasis of neuroendocrine tumour for serotonin, small cell carcinoma for CD56 and NSE, and colonic mucosa for chromogranin A and synaptophysin. PBS instead of primary antibody was used as a negative control. Immunohistochemical evaluation of NE markers Cytoplasmic immunostaining for serotonin, CD56, NSE, chromogranin A and synaptophysin was classified dichotomously as positive or negative simultaneously by three observers (HR, PH and SK). Statistical analyses SPSS for Windows 15 (Chicago, IL, USA) was used for statistical analyses. Statistical significance between stainings and clinicopathological parameters was determined using the chi-squared test or Fisher's exact test in the case of low expected frequencies. Corrected cancer-specific survival was analysed with the Kaplan-Meier curve and the significance with the log rank test. The Cox regression model was used for multivariate analysis. Results Patients, follow-up and treatment The median age of the patients was 63 (range 29-86) years. Seventy-seven (51%) patients were women, 75 (49%) men. Seven tumours (5%) were resected and 145 (95%) were operated by radical nephrectomy. The median follow-up time was 90 (range 0-209) months and followup was complete in all cases. During follow-up, 44 (29%) patients died of RCC, 40 (26%) died of other causes and 68 (45%) patients were still alive. The distribution of clinicopathological parameters of the tumours is described in Table 1. Immunohistochemical findings Twelve tumours (8%) were stained for serotonin and 26 (18%) for CD56. The immunopositivity for NSE was detected in 69 cases (48%). The immunopositivity for serotonin and CD56 were detected in the same tumours (p < 0.001). All tumours were immunonegative for chromogranin A (100%). Two of the tumours (1%) were immunopositive for synaptophysin. The immunopositivity for NSE was more common in clear cell RCCs than other subtypes (p = 0.01). There was Ronkainen et al. Diagnostic Pathology 2010, 5:28 http://www.diagnosticpathology.org/content/5/1/28 Page 3 of 7 no association between the immunoreactivity for other NE markers and histological subtype of the tumours (Table 2). The RCC-specific survival of patients with serotonin positive tumours (Figure 1) as well as CD56 and NSE immunonegative tumours was somewhat better but the difference was not significant (Table 3). The two patients with synaptophysin immunopositive tumours showed an excellent RCC-specific survival. Immunostaining for serotonin, CD56 and NSE did not associate with grade or stage (Table 4). Each staining (serotonin, CD56 and NSE) was separately included in the Cox multivariate analysis with stage, Fuhrman grade and age. The only significant prognostic factor for RCCspecific survival was stage (p < 0.001). Discussion Evidence regarding the pr