Human Pathology (2012) 43, 150–152
www.elsevier.com/locate/humpath
Correspondence Necrosis assessment in renal carcinoma To the Editor: We have read with interest the article by Dr Si et al [1] reporting the clinical and pathologic features of 14 renal small cell carcinoma from 2 institutions. The tumors had a mean size of 7.1 cm and were extremely aggressive, with perinephric adipose tissue invasion at the time of diagnosis in 13 of the 14 patients. Interestingly, tumor necrosis was observed in all cases and ranged from 10% to 50% of the tumor area. As underlined by the authors, renal small cell carcinoma should be differentiated from the other renal neuroendocrine tumors. Renal carcinoid tumors have a low mitotic rate and lack necrosis [2]. Large cell neuroendocrine carcinomas have a high mitotic rate, at least 1 neuroendocrine marker, and necrosis. Hence, tumor necrosis is important to characterize kidney cancers. In addition, tumor necrosis is correlated with VEGF (vascular endothelial growth factor) expression and is an important indicator to evaluate kidney cancer response to the new anti–VEGF-targeted therapies [3]. However, tumor size changes obtained with computed tomography (CT) or magnetic resonance imaging are usually delayed [4]. New functional imaging techniques could be able to assess the tumor response earlier than pathologic analysis, but these techniques are either difficult to apply to patients (long breath-hold required for dynamic contrast enhancement analysis) or not widely used in radiology [4]. We assessed tumor necrosis area in surgical pieces of 41 patients with renal cell carcinoma. Whole tumor sections were scanned at original magnification ×100 using an automated digital microscope. On digitized slides stained with hematoxylin-eosin, the tumor area analyzed was proportional to the largest dimension of each tumor. Necrotic and non-necrotic tumor areas were directly selected on screen and measured by the image analysis software. The same observers also semiquantitatively measured tumor necrosis on the same tissue sections using conventional microscopy. Necrotic areas ranged from 0% to 90% (median, 25%) of the whole tumor area. Interobserver agreement for necrosis extension was good, with digital microscopy (intraclass correlation coefficient, 0.96) higher than with conventional microscopy (intraclass correlation coefficient, 0.75). We then compared the results with those obtained, before surgery, on contrast-enhanced CT focused on the tumor. 0046-8177/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
Four radiologists independently evaluated the tumor necrosis area versus the tumor area ratio on CT. We showed that tumor necrosis evaluation was not reproducible between the 4 different radiologists based on CT data (concordance coefficient, 0.714; 95% confidence interval, 0.32-0.76). In addition, there was no correlation between imaging and pathology data on the surgical pieces (correlation coefficient, 0.22; 95% confidence interval, −0.27 to 0.62). Therefore, for kidney cancers where necrosis assessment is necessary for tumor characterization as well as for evaluation of response to antiangiogenic targeted therapies, pathology analysis of whole tumor specimen remains irreplaceable.
Cédric de Bazelaire Université Paris Diderot Sorbonne Paris Cité F-75010 Paris, France INSERM, U728-Paris F-75010 France AP-HP-Hôpital Saint-Louis Service de Radiologie Paris, F-75010 France E-mail address:
[email protected] Jacqueline Rivet Mariana Varna Université Paris Diderot Sorbonne Paris Cité F-75010 Paris, France INSERM, U728-Paris F-75010 France AP-HP-Hôpital Saint-Louis Laboratoire de Pathologie Paris, F-75010 France E-mail addresses:
[email protected] [email protected] Sandrine Katsahian Université Paris Diderot Sorbonne Paris Cité F-75010 Paris, France Inserm U717-Paris F-75010 France E-mail address:
[email protected]
Correspondence
151
Christophe Leboeuf Université Paris Diderot Sorbonne Paris Cité F-75010 Paris, France INSERM, U728-Paris F-75010 France AP-HP-Hôpital Saint-Louis Laboratoire de Pathologie Paris, F-75010 France E-mail address:
[email protected] Wissam Sandid AP-HP-Hôpital Saint-Louis Laboratoire de Pathologie Paris, F-75010 France E-mail address:
[email protected] Philippe Bertheau Anne Janin Université Paris Diderot Sorbonne Paris Cité F-75010 Paris, France INSERM, U728-Paris F-75010 France AP-HP-Hôpital Saint-Louis Laboratoire de Pathologie Paris, F-75010 France E-mail addresses:
[email protected] [email protected]
doi:10.1016/j.humpath.2011.09.005
References [1] Si Q, Dancer J, Stanton ML, et al. Small cell carcinoma of the kidney: a clinicopathologic study of 14 cases. HUM PATHOL 2011;42:1792-8. [2] Lane BR, Chery F, Jour G, et al. Renal neuroendocrine tumours: a clinicopathological study. BJU Int 2007;100:1030-5. [3] Rioux-Leclercq N, Fergelot P, Zerrouki S, et al. Plasma level and tissue expression of vascular endothelial growth factor in renal cell carcinoma: a prospective study of 50 cases. HUM PATHOL 2007;38:1489-95. [4] de Bazelaire C, Alsop DC, George D, et al. Magnetic resonance imaging-measured blood flow change after antiangiogenic therapy with PTK787/ZK 222584 correlates with clinical outcome in metastatic renal cell carcinoma. Clin Cancer Res 2008;14:5548-54.
Necrosis assessment in renal carcinoma—reply In reply: We thank de Bazelaire et al for their interest in our article [1] and appreciate the opportunity to discuss the role of tumor necrosis in renal cell carcinoma. Tumor necrosis is often present in renal cell carcinomas. It is generally believed that tumor necrosis occurs when the blood supply cannot meet a tumor's rapid growth,
reflecting aggressive biologic behavior. Several studies have found that the presence of tumor necrosis in renal cell carcinoma is closely associated with high tumor stage and poor prognosis [2-3]. However, there is also conflicting evidence on whether tumor necrosis can provide any additional prognostic information beyond that of tumor grade and stage. Recent studies have found that the extent of tumor necrosis may provide more accurate prognostic information than simple presence or absence of tumor necrosis [4]. A 3-tiered reporting system—no necrosis, lower necrosis (≤20%), and high necrosis (N20%)—has been proposed and appears to have appropriate predictive accuracy [4]. However, it is difficult to accurately assess the extent of tumor necrosis in renal cell carcinoma. Tumor necrosis in renal cell carcinoma can be evaluated by several methods, including radiographic, macroscopic, and microscopic examination. Radiographic techniques, including computed topography, may identify increasingly small tumors in the kidney but is generally not effective for detecting tumor necrosis, a pathologic process. Tumor necrosis can be assessed on gross examination, but it should be differentiated from fibrosis, hyalinization, cystic degeneration, and hemorrhage, often necessitating microscopic verification. In most studies in the literature, including ours, the extent of tumor necrosis is evaluated by microscopic examination of the renal tumor. However, microscopic examination can hardly be regarded as a perfect way of defining the extent of tumor necrosis because it involves counting only microscopic foci of necrosis on a limited number of histologic slides. Because the microscopic slides sample a distinctively small portion of the tumor, the extent of tumor necrosis on the slides may well be affected by where the sections were taken from the tumor. Our study suggests that small cell carcinoma of the kidney has a closer relationship with urothelial carcinoma (arising from the renal pelvis and calyces) than renal cell carcinoma (arising from the renal tubules). We estimated the percentage of tumor necrosis by comparing the coagulative necrotic tumor area to the entire tumor area under a conventional microscope, which was subject to interobserver variation. de Bazelaire et al reported their experience of assessing the extent of tumor necrosis in renal cell carcinoma using an automated digital microscope. Although this method showed better reproducibility than that using a conventional microscope, it may not result in improved accuracy. They stated that the tumor area was proportional to the largest dimension of the tumor, which is difficult to apply to a microscopic examination. Coagulative necrosis is not only irregular in shape but also multifocal and intermixed with viable tumor cells in renal cell carcinomas. With these considerations in mind, we think that more accurate assessment of the extent of tumor necrosis in renal cell carcinoma can be achieved with a balanced approach combining macroscopic and microscopic examination.
