Background: Papillary renal cell tumours (RCTs) have been described as a genetic entity. Recently, papillary RCTs have been divided into small (type 1) and large (type 2) cell tumours. Subsequent DNA analyses have resulted in controversial data regarding putative genetic changes marking type 1 and type 2 tumours.
Aim: The aim of this study was to improve the original description that papillary RCT is a genetic entity regardless of the phenotypic variation.
Methods: DNA from 163 papillary RCTs, including 82 multiplex tumours from eight hereditary cases, was analysed for copy number changes by chromosomal comparative genomic hybridisation (CGH) and/or for allelic changes at chromosomes 7 and 17 by microsatellite analysis. The results of the genetic analysis were compared with the cytological characteristics of the tumours.
Results: The results showed alterations of chromosomes 7 and 17 at similar frequencies in papillary RCTs with characteristics ranging from small to large cell, nuclear grade 1 to 3, and 3 mm to 16 cm diameter.
Conclusion: Trisomies of chromosomes 7 and 17 are specific genetic alterations in papillary RCTs irrespective of their size, grade and cellular differentiation.
Statistics from Altmetric.com
Papillary renal cell tumours (RCTs) displaying characteristic genetic changes of trisomies of chromosomes 7, 8, 12, 16, 17 and 20 were identified as a genetically well-defined entity nearly two decades ago1 ,2; in the first report, a variation of cytological pattern from small basophilic cuboidal to large eosinophilic columnar cells was presented.1 Following the Expert Meeting on Classification of Renal Cell Tumours held in Heidelberg,3 two of the participants classified RCTs strictly into small cell (type 1) and large cell (type 2) papillary RCTs.4 Subsequently, several authors described genetic changes supporting the cytological subtyping,5 ,6 which is now fixed in the World Health Organization (WHO) blue book series of tumour classification.7
It is generally believed that papillary RCTs develop from the adult renal tubular system. However, based on the association of nephrogenic-rest-like lesions and clinically recognised cancers it has been suggested that papillary RCTs develop from partially differentiated nephrogenic rests, and therefore that the cellular characteristics of tumours correspond to the level of epithelial differentiation of precursor lesions.8 ,9 If the hypothesis is correct that papillary RCTs develop from small precursor lesions carrying trisomy 7 and 17, then the alteration of these chromosomes should occur in all papillary RCTs irrespective of the cellular phenotype. To further investigate this hypothesis we analysed 163 papillary RCTs including 82 multiplex tumours from eight hereditary cases with germline mutation of MET for alterations of chromosomes 7 and 17 by microsatellite analysis and/or chromosomal comparative genomic hybridisation (CGH).
Materials and methods
Tissue samples and cytological evaluation
Fresh tumour and normal tissues were obtained immediately after nephrectomy from the Department of Urology, Hannover Medical School, and the Department of Urology, Ruprecht-Karls University of Heidelberg, Germany. We also analysed additional tumours from paraffin-embedded tissues obtained from the Department of Urology, University of Padova, Italy (hereditary cases with germ line mutation10) and the Department of Urology, University of Umeå, Sweden. Histological diagnosis was established according to the Heidelberg Classification of Renal Cell Tumours and WHO.3 ,7 The size of tumours varied between 3 mm and 16 cm in diameter. Grading was performed based on the three grade scales, where G1 and G2 corresponds to Fuhrmann G1 and G2, whereas G3 corresponds to Fuhrmann G3–G4. A part of this study was performed on routinely collected tumour samples, the use of which for scientific purposes was in accordance with ethical guidelines of the University of Padova and the University of Umeå at that time. The use of all tumour specimens for this study was approved by the ethics committee of the University of Heidelberg.
To demonstrate the variation in cell morphology and the stability of genetic changes we evaluated all papillary RCTs based on the cytological features according to the examples shown in fig 1A–C. Tumours composed entirely of small cuboidal cells with scanty “clear” or eosinophilic cytoplasm resembling blastemal cells were called small basophilic (SB) or small eosinophilic (SE) cell tumours, respectively. Tumours with large columnar cells with abundant granular–eosinophilic but sometimes pale eosinophilic cytoplasm were designated as large eosinophilic (LE) cell tumours. Tumours composed of cells in size between the small and large cells were called medium cell tumours. The cells of medium size tumours displayed eosinophilic (ME), basophilic (MB) or clear (MC) cytoplasm. The remaining tumours were split into two additional groups, namely those consisting of a mixture of small and medium cells or a mixture of medium and large cells. However, it should be acknowledged that in a substantial number of cases there is no strict border between small, medium or large cells with clear, basophilic or eosinophilic cytoplasm, but there is a continuity in cell size and staining intensity.
Because most papillary RCTs are heavily infiltrated by lymphocytes/plasmocytes and foamy cells, DNA was isolated from short term cultures of papillary RCTs and from corresponding normal kidney tissues. When only paraffin blocks were available, one H&E-stained section of 4 μm was used to determine the most appropriate part of tumour containing the lowest number of stromal or foamy cells. This part of tumour was then microdissected from subsequent unstained sections of 8 μm thickness. The slices were deparaffinised with xylene treatment and rehydrated in an alcohol series. DNA was extracted from cell cultures and paraffin materials by phenol/chloroform precipitation after proteinase K digestion.
Comparative genomic hybridisation
CGH analysis of 45 papillary RCTs was performed as described previously, with minor modifications.10 ,11 Image acquisition, processing and evaluation were performed using a Leica DM RXA RF8 epifluorescence microscope (Leica, Bensheim, Germany) equipped with a Sensys CCD camera (Photometrics, Tucson, Arizona, USA; Kodak KAF 1400 chip) controlled by Leica Q-FISH software (Leica Microsystems Imaging Solutions Ltd, Cambridge, UK). Three colour images were acquired from 10 metaphases per sample and these were processed using the Leica Q-CGH software. The threshold values for detection of genomic imbalances were 0.75 for losses and 1.25 for gains.
For analysis the short and long arms of chromosomes 7 and 17, the following microsatellites were used: D7S817, D7S793, D7S1791 (7p); D7S1824, D7S2847, D7S1820 (7q); D17S969, D17S1298, D17S1537 (17p); D17S1795, D17S1824, D17S787, D17S1306 (17q). The microsatellite analysis was carried out until one or two loci at each chromosomal arm was informative of the allelic changes, as described previously.12 The results were evaluated according to the following scoring system: when normal as well as tumour DNA revealed the same pattern, it was scored as retention. When one allele in tumour DNA was reduced to approximately 50% of that seen in normal DNA, it was scored as allelic imbalance. When one allele was not detected in tumour DNA or only a slight signal was seen (especially in cases of dinucleotide repeats), it was scored as loss of heterozygosity.
Results of cytomorphological evaluation of 163 papillary RCTs are shown in table 1. Some examples are shown in fig 1. A solid growth of small “blue cells” resembling nephrogenic rests with glomeruloid structures was seen in 14 of the 72 SB cell tumours. Most small cell or sometimes medium or mixed small–medium cell areas of the tumours showed a solid or solid–tubular growth pattern or papillary structures. The growth characteristic of tumours designed as medium cell tumours varied from solid through tubular to papillary structures, with tubulo-papillary being dominant in most cases. Typical papillary and tubulopapillary structures were seen in all tumours composed of large eosinophilic cells. Nuclear pseudostratification was seen in most of the large cell tumours and also in several tumours displaying medium size cells.
The cytoplasm of tumour cells showed a variation from nearly empty through pale towards strong eosinophilic staining intensity. Some of the tumours with medium size cells displayed basophilic cytoplasmic staining. The staining intensity, which simply reflects the number of mitochondria and ribosomes, was variable even within the same tumour. The group of SE/SB tumours displayed nuclear grade 1 or 2, the ME/MB/MC tumours showed G2 and G3/G4 nuclear grade, whereas large cell tumours displayed G2 and G3/G4 nuclear grade. Similarly to the variation in cytoplasmic staining and cell size, the nuclear grade varied within the same tumour in some cases.
Alteration of chromosomes 7 and 17
Chromosomal CGH analysis of high molecular weight DNA from 45 papillary RCTs revealed a combined trisomy of chromosomes 7 and 17 in 38 cases; five cases showed gain only at chromosome 17, one case only at chromosome 7, and in one papillary RCT only duplication of chromosome 1q was seen. None of the papillary RCTs showed copy number loss at chromosome 7 or 17. Therefore, an allelic imbalance observed by microsatellite analysis was evaluated as duplication of one allele (fig 2). An allelic imbalance (allelic duplication) at both chromosomes or one of the two chromosomes occurred in 98% of the papillary RCTs. Allelic duplications corresponding to combination of trisomy of chromosomes 7p/7q and 17p/17q was noticed in all cellular variants (table 1). No differences were seen when evaluating hereditary versus sporadic papillary RCTs.
By evaluating each chromosomal arm individually, it was found that they were involved in allelic changes in 85–91% of the cases, with allelic duplication at chromosome 17q (91%) accounting for the highest number of changes (table 2). The frequency of genetic changes was similar in each group of small, medium and large cell tumours.
In the present study we detected alterations of chromosomes 7 and 17 at similar frequency in papillary RCTs with small to large cell characteristics, nuclear grade 1 to 3/4, and 3 mm to 16 cm diameter. Small papillary precursor lesions/adenomas analysed from hereditary and sporadic cases have been shown to display trisomy 7 and 17 but no other autosomal change.2 ,13 Therefore, trisomy 7 and 17 may account for the primary genetic change occurring in the early stage of tumour development. In two comprehensive histological studies investigating entire nephrectomy specimens with sporadic and hereditary papillary RCTs, averages of 42 and over 1000 papillary lesions has been found per kidneys, respectively.8 ,14 These data support our hypothesis that papillary RCT develops from precursor lesions carrying trisomy 7 and 17 and that the cellular phenotype of tumours displays a continuous spectrum of cellular differentiation from small basophilic blastem-like cells towards large eosinophilic columnar cells.
A comprehensive study on chromosomal alteration and phenotype described trisomy 7 and 17 at similar frequency in all cytomorphological variations of RCTs and suggested that type 2 papillary RCTs evolve from type 1 tumours.15 Two other studies also excluded trisomy 7 and 17 as a marker for low-grade versus high-grade or basophilic versus eosinophilic cell types.16 ,17 In contrast, other studies found that genetic changes such as gains of chromosomes 7p and 17p or allelic imbalance at 17q and 9p differentiated type 1 from type 2 tumours.5 ,6 In another study loss of chromosome 3p, which is highly characteristic for conventional RCTs, was found in 50% of chromophilic RCTs referred to as papillary RCT.18 These data suggest that several granular/eosinophilic conventional RCTs with papillary growth pattern might have been included in those series, resulting in conflicting data regarding histological characteristics, immunohistochemical parameters and outcome of the disease.17 ,19 ,20 ,21
There are two major problems with studies transferring new knowledge on the genetics of RCTs upon the cytomorphological classification systems. First, that chromophilic (granular) cellular phenotype does not correspond to papillary RCTs. It was stated by van den Berg et al that “although trisomy of chromosomes 7 and 17 is more frequent”, “it is not specific for chromophilic RCT per se”.22 Eble and Delahunt have also acknowledged that “the majority of tumors in both groups were classified as grade 2, which would indicate that if tumours were divided according to nuclear grade alone, then 68% of type 1 and 60% of type 2 papillary renal cell carcinoma in this series would have been assigned to the same group”.4 Another study has also classified type 1 and type 2 papillary RCTs, but designated 24% of the tumours as mixed cell type.23 A comprehensive study on the global gene expression of papillary RCTs has separated two molecular subclasses with distinct survival probability and showed that tumours with high survival probability include small, small/large mixed and also large cell tumours.24 Based on these data and our previous and present studies, we do not propose to divide papillary RCTs into any cytological subtypes. In this study, we simply used a complex cytomorphological evaluation to demonstrate the continuity in cellular differentiation without strict border between any cell types and also to demonstrate the occurrence of trisomy of chromosomes 7 and 17 throughout all papillary RCTs with cellular variation.
In conclusion, trisomy of chromosomes 7 and 17 is a specific genetic alteration in papillary RCTs irrespectively of their size, grade and cellular differentiation. Papillary RCTs display the whole spectrum of cellular differentiation from small blastem-like “blue” cells to large differentiated epithelial cells. Most of the clinically recognised tumours consist of medium size cells leading to high intraobserver and interobserver variation in dividing papillary RCTs strictly into type 1 and type 2 tumours. This may also explain the controversial results obtained by clinicopathological studies. Therefore, we propose that (1) papillary RCTs with characteristic DNA alterations should be diagnosed as an entity and the cellular and growth characteristics, without any subtyping, should only be included into the description; (2) all clinical studies aiming to estimate the prognosis as well as cDNA array analysis or immunohistochemical studies should be carried out on series of papillary RCTs diagnosed by the occurrence of specific DNA alterations.
Papillary renal cell tumours (RCTs) are characterised by trisomy of chromosomes 7 and 17.
Trisomy of chromosomes 7 and 17 occurs in all cytological variations of papillary RCTs at similar frequency irrespective of grade or size.
The cytomorphology of papillary RCTs is variable within the same tumour, whereas the genetic changes are constant.
The authors thank Drs Prayer-Galetti and Börje Ljungberg for providing paraffin blocks from hereditary papillary RCTs.
Funding This work was supported by the Deutsche Forschungsgemeinschaft (Ko 841/30-1).
Competing interests None.
Ethics approval Ethics approval was obtained from the ethics committee of the University of Heidelberg.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.