Custom Search

Significance of Lymphatic Invasion and Proliferation on Regional Lymph Node Metastasis in Renal Cell Carcinoma

Abstract and Introduction

Abstract

We studied the associations of lymphatic invasion and lymphatic vessel density around tumors with lymph node (LN) status in renal cell carcinoma (RCC) by immunohistochemical analysis using D2-40 antibody as a lymphatic marker. Surgically removed specimens from 76 cases with RCC, including 16 cases with LN metastasis, were used. Lymphatic vessel density around the tumor increased compared with normal kidneys but was not significant by LN status. Tumor size, tumor cell types, patterns of tumor growth, nuclear grade of tumor cells, venous invasion, lymphatic invasion, and primary tumor stage were predictive factors for LN metastasis. Based on multivariate regression analysis, only lymphatic invasion was an independent risk factor for LN metastasis. The immunohistochemical detection of lymphatics was useful for identifying the lymphatic invasion of RCC, and the presence of lymphatic invasion around RCC was an independent predictive factor for LN metastasis.

Introduction

Various prognostic factors for renal cell carcinoma (RCC) have been clinically, histologically, or genetically examined,[1] and the most important factor is considered to be tumor stage.[2-4] In the TNM clinical classification of RCC,[5] tumor stage has been established based on 3 factors: the primary tumor (pT) stage, including the size and the degree of tumor extension, lymph node (LN) metastasis, and distant metastasis. The presence of LN metastasis is, thus, an important factor for determining the clinical stage and for predicting patient outcome.[6,7]

The role of an extensive lymphadenectomy in the surgical management of RCC remains controversial.[8,9] Several reports have indicated that a radical nephrectomy with extensive lymphadenectomy does not affect patient survival.[10-13] However, divergent opinions have been reported for lymphadenectomy in patients with RCC; for example, a systematic lymphadenectomy should be performed because of an increased survival in patients with positive LNs[14,15] and the necessity for accurately determining the tumor stage.[16] Because the former has recently been the dominant opinion, neither a systematic nor an ipsilateral lymphadenectomy has been performed for patients with RCC who did not have clinically evident lymphadenopathy before or during surgery. In such RCC cases, the LN status by histopathologic examination remained unclear.

Metastasis of RCC to the regional LNs occurs through cancer cell invasion into the lymphatic route from its beginning in the cortical interstitium to the large-caliber lymphatics associated with blood vessels.[17] The recent development of antibodies against lymphatic endothelial cell–specific proteins, such as lymphatic endothelial hyaluronan receptor-1, podoplanin, prox- 1, and D2-40, has made it possible to discriminate lymphatics from blood vessels in tissue sections examined by immunohistochemical analysis.[18,19] Our recent report showed that the lymphatic distribution in renal tissues with normal and pathologic conditions could be clearly detected by immunohistochemical staining using the D2-40 antibody.[20] In addition, the number of lymphatics distributed in the cortex around RCC significantly increased in comparison with the normal cortex, and small lymphatic vessels invaded by RCC cells were rarely observed at the tumor margin or in the vicinity of the tumor.[20] However, no histopathologic report has previously demonstrated the significance of lymphatic density around RCCs or lymphatic invasion of RCC cells on metastasis to regional LNs.

In the present study, we carried out a histopathologic investigation using D2-40 immunohistochemical staining to elucidate the associations of lymphatic invasion of RCC cells and lymphatic density around the tumor with LN metastasis through a comparison between 60 RCC cases with no LN metastasis and 16 RCC cases with LN metastasis.

Materials and Methods

Cases and Samples

A total of 76 cases were used for this study. We examined nephrectomy tissue samples for RCC obtained from January 1995 to October 2004 at Toho University Medical Center Ohmori Hospital, Tokyo; National Defense Medical College, Tokorozawa; and Saiseikai Kanagawa-ken Hospital, Yokohama, Japan, and all patients underwent regional LN resection of more than 1 LN during a nephrectomy Table 1 . Lymphadenectomy included ipsilateral removal of the renal hilar, paracaval, and/or aortocaval LN(s). When the swelling of contralateral LNs was found before or at the time of surgery, the contralateral LNs were also removed. Written informed consent to use operatively removed renal tissue samples was provided by patients at each medical institute, and the present study was approved by the committee on ethical standards of Toho University School of Medicine.

Renal tissue samples were fixed with 10% neutral buffered formalin and embedded in paraffin. All cases included the tumor parts and normal cortices around the tumor. They were divided into 2 groups according to LN status by histopathologic examination of the LNs with H&E staining: (1) LN+, with metastasis to regional LNs, 16 cases; and (2) LN–, without LN metastasis, 60 cases. Tumor size was determined by a review of the gross description, and the longest axis was defined as tumor size at the maximal cut-surface area. The number of sampled LNs showed no significant difference between LN+ and LN– groups.

We used 5 autopsied cases as normal control cases for the examination of lymphatic distribution in the renal cortex. Patients had died of causes other than RCC or renal failure and had no significant lesion in the kidneys except for mild arteriosclerosis or congestion. One piece of normal renal tissue per kidney was dissected, and all 10 renal tissue samples were fixed with 10% neutral buffered formalin. Thin sections were used for H&E–Victoria blue staining and D2-40 immunohistochemical staining.

Histopathologic Examination

By observation of H&E-stained sections, the histologic type of each RCC case was evaluated according to the 2004 World Health Organization (WHO) histologic classification of tumors of the kidney,[21] and RCC cases were thus divided into 4 categories: (1) clear cell RCC, (2) papillary RCC, (3) multilocular clear cell RCC, and (4) spindle cell carcinoma, because RCC cases of the other histologic type were not found in our cases. RCC cases were also divided into 5 categories according to the cell types composing the RCC: (1) clear, (2) granular, (3) mixed, (4) pleomorphic, and (5) spindle. In this study, clear cell–type RCC was defined as having tumor cells exhibiting clear cytoplasm by H&E staining; granular cell–type RCC is a specific tumor formerly categorized as granular cell carcinoma in the 1998 WHO classification of kidney tumors;[22] pleomorphic cell–type RCC is composed of various-sized tumor cells that have pleomorphic nuclei and sometimes exhibit giant cells; spindle cell–type RCC is also a specific tumor, which is composed of spindle-shaped tumor cells and almost synonymous with spindle cell carcinoma in the 2004 WHO classification of kidney tumors;[21] mixed type exhibits several cell types of at least 2 cell types, in which 1 cell type constitutes at least 30% of the sample. Histologic and cell types used in this study were predominant types that constituted more than 80% of the tumor.

The nuclear grade of RCC cells was determined by observations under high-power microscopy (×400) according to the criteria of Fuhrman et al.[23] Grade 1 tumors were composed of cells with round and uniform nuclei that were larger than normal nuclei of tubular epithelial cells. Grade 2 tumor cells had larger nuclei than those in grade 1 tumor cells, and the nuclei exhibited irregularities in outline and included nucleoli. Grade 3 tumor cells had even larger nuclei than grade 2 tumor cells, possessing an obviously irregular outline and prominent large nucleoli. Grade 4 tumor cells had bizarre, often multilobed nuclei with heavy chromatin clumps.

RCC cases were divided into 2 categories according to the pattern of tumor growth: expanding type and infiltrating type. The expanding type had a fibrous capsule or fibrous lesion around the tumor, which distinctly bordered on the nontumor renal parenchyma, and cancer cells did not invade into the fibrous region. The infiltrating type had an indistinct border with nontumor renal parenchyma in which cancer cells invaded the interstitium. In addition, RCC cases were divided into 2 groups by the existence of venous invasion, which was histologically defined by H&E–Victoria blue staining and macroscopic observation at surgery. RCC cases showing tumor invasion into the inferior vena cava, sinus veins, and/or intrarenal vein were defined as venous invasion–positive cases.

For accurate determination of the pT stage, perinephric tissues including renal sinus fat, sinus veins, Gerota fascia, and adrenal gland were histologically observed. From such histologic examination in addition to macroscopic observation at surgery, pT stage in each case was determined according to the TNM classification.[5]

Immunohistochemical Staining

All thin sections of renal tissues (>3 sections from each case) were immunostained with antibodies against D2-40 (1:50; Signet Laboratories, Dedham, MA), CD68 (1:80; DakoCytomation, Carpinteria, CA), and vascular endothelial growth factor-C (VEGF-C; 1:75; Zymed, South San Francisco, CA). Before immunostaining, all sections were pretreated with 10 mmol/L of citrate buffer for 10 minutes at 95°C for antigen retrieval. Immunohistochemical staining was performed by using the CSA II kit (DakoCytomation) according to the manufacturer's instructions. The sections were visualized by treating the slides with diaminobenzidine tetrahydrochloride. To verify antibody specificity, sections from each paraffin block were used as negative control samples by replacing the primary antibody with normal mouse or rabbit immunoglobulin.

In RCC cases, the sections immunostained for D2-40 were observed by light microscopy at ×100 magnification. The number of lymphatics in a ×100 field of view of the cortex adjacent to the tumor was counted at random in 20 fields, and the average number of lymphatics in the ×100 field of view was defined as lymphatic density around the tumor. Sections immunostained for VEGF-C were also observed by light microscopy, and RCC cases with positive tumor cells at the rim of the tumor were defined as VEGF-C+. The number of lymphatics invaded by RCC cells in the peripheral region of the tumor and in the interstitium around the tumor was counted on all tissues of the cut surface at the largest diameter of the tumor.

In the normal renal sections immunostained for D2-40, the number of D2-40+ lymphatics in the cortex was counted at random in the same way as in RCC cases, and the average number of lymphatics in a ×100 field of view was defined as lymphatic density of the normal cortex.

Statistical Analyses

Statistical analyses were performed using the χ2 test, Fisher exact test, Scheffe F test, and Mann-Whitney U test to assess the significance of the impact of each subset of histopathologic factors on LN status and lymphatic invasion. A difference at a P value of less than .05 was considered statistically significant. The StatView program (SAS Institute, Raleigh, NC) was used for all analyses. In addition, multivariate logistic regression analyses of lymphatic invasion and LN status were performed by using a multilogistic model (Excel Life Span Model, Esumi, Tokyo, Japan). For multivariate logistic regression analyses, RCC cases were divided into 2 groups according to nuclear grade: (1) nuclear grades 1 and 2 and (2) nuclear grades 3 and 4.